Search

While laboratory methods of elemental analysis of soil nutrients are used frequently to support soil studies, the implementation of more portable and cost-efficient methods lingers behind. The portable (handheld) X-ray fluorescence spectrometer (XRF) is one such tool enabling onsite elemental analysis in a straightforward manner. However, in soil studies the use of XRF often remains cumbersome, following the poor performance of the method for low-Z elemental analysis and the complex nature of the soil matrix, introducing background noise. Here, we therefore evaluate how the potential use of a portable XRF for predicting potassium (K), phosphorus (P), magnesium (Mg) and calcium (Ca) can be improved through the analysis of XRF spectral data with the Random Forest (RF) machine learning method. A total of 105 soil samples from a wide range of soils collected from 10 different countries (D.R. Congo, Belgium, Ivory Coast, Italy, The Netherlands, Saudi-Arabia, South Africa, Spain, Switzerland, and Zimbabwe) were scanned using an Oxford XMET8000 XRF spectrometer (Oxford Instruments, UK). Spectral data of the calibration set (n = 74) were pulled in one matrix alongside measured elemental concentrations and subjected to RF analysis. Resulting models were validated using an independent validation set (n = 31). The best RF prediction result was obtained for K followed successively by Ca, Mg and P with coefficient of determination (R2) values of 0.83, 0.76, 0.69, and 0.47, and root mean squared error of prediction (RMSEP) of 2283.8, 6818.7, 1511.8, and 538.08 mg kg−1, respectively. The RF modelling procedure provided improved prediction performance compared to the calibration models provided by the manufacturer (R2 = 0.65, 0.75, 0.65, and 0.22, for K, Ca, Mg and P, respectively). Our results suggest that portable XRF instruments coupled with spectral data analysis by RF allows for rapid and low-cost analysis of soil K and Ca with remarkable accuracy. Still, the lower measurement accuracy for P and Mg suggests further work is needed to test whether the prediction can be improved by better calibration models, and how such approaches can help overcoming instrumental limitations.

Land use change can significantly influence both mineralogy and chemical soil properties. This conversion, particularly from forest to agricultural system occurs often in volcanic soils due to their favorable properties for food production. Under agriculture, minerals can weather faster than in natural vegetation and this also impacts soil functioning. We aim to assess the impact of land use on geochemical soil properties and soil organic carbon across soils of different age. This study was conducted in Mt. Tangkuban Perahu and Mt. Burangrang where the soils were derived from similar andesitic parent material and have different ages based on their lithology. Five sites were selected representing land uses that have been converted (pine forest and agricultural land) and one site of natural forest as the origin of land use. The results showed that land use management enhances the mineral transformation. Pine forest and agricultural sites displayed higher weathering degree than natural forests as indicated by higher clay content, iron crystallinity index and the presence of gibbsite. The weathering degree of soils in agricultural sites might result from the length of cultivation period and soil age. Land use conversion also altered chemical properties such as pH, CEC, basic cations, and the proportion of amorphous materials. Non-crystalline Al and Fe minerals as indicated by Alo+1/2Feo were highly correlated with organic carbon and specific surface area (SSA) in the subsoils of all land uses. However, we did not see the accumulation of organic carbon in subsoils compared to topsoils as the amount of non-crystalline Al and Fe minerals increases with depth, especially in agricultural lands where the organic fertilizer input is very high. In addition, a significant proportion of carbon was stored in sand aggregate fractions in agricultural land which have longer cultivation period, while it was more readily found in silt and clay fractions in the site with shorter period.

Quantifying soil carbon dynamics is of utmost relevance in the context of global change because soils play an important role in land-atmosphere gas exchange. Our current understanding of both present and future carbon dynamics is limited because we fail to accurately represent soil processes across temporal and spatial scales, partly because of the paucity of data on the relative importance and hierarchical relationships between microbial, geochemical and climatic controls. Here, using observations from a 3,000-kyr-old soil chronosequence preserved in alluvial terrace deposits of the Merced River, California, we show how soil carbon dynamics are driven by the relationship between short-term biotic responses and long-term mineral weathering. We link temperature sensitivity of heterotrophic respiration to biogeochemical soil properties through their relationship with microbial activity and community composition. We found that soil mineralogy, and in particular changes in mineral reactivity and resulting nutrient availability, impacts the response of heterotrophic soil respiration to warming by altering carbon inputs, carbon stabilization, microbial community composition and extracellular enzyme activity. We demonstrate that biogeochemical alteration of the soil matrix (and not short-term warming) controls the composition of microbial communities and strategies to metabolize nutrients. More specifically, weathering first increases and then reduces nutrient availability and retention, as well as the potential of soils to stabilize carbon.

International audience; To protect soils against threats, it is necessary to predict the consequences of human activities and global change on their evolution on a ten to hundred year time scale. Mechanistic modelling of soil evolution is then a useful tool. We analysed the ability of the SoilGen model to be used for projections of soil characteristics associated to various soil threats: vertical distributions of

To accurately estimate past terrestrial carbon pools is the key to understanding the global carbon cycle and its relationship with the climate system. SoilGen2 is a useful tool to obtain aspects of soil properties (including carbon content) by simulating soil formation processes; thus it offers an opportunity for both past soil carbon pool reconstruction and future carbon pool prediction. In order to apply it to various environmental conditions, parameters related to carbon cycle process in SoilGen2 are calibrated based on six soil pedons from two typical loess deposition regions (Belgium and China). Sensitivity analysis using the Morris method shows that decomposition rate of humus (kHUM), fraction of incoming plant material as leaf litter (frecto) and decomposition rate of resistant plant material (kRPM) are the three most sensitive parameters that would cause the greatest uncertainty in simulated change of soil organic carbon in both regions. According to the principle of minimizing the difference between simulated and measured organic carbon by comparing quality indices, the suited values of kHUM, (frecto and kRPM in the model are deduced step by step and validated for independent soil pedons. The difference of calibrated parameters between Belgium and China may be attributed to their different vegetation types and climate conditions. This calibrated model allows more accurate simulation of carbon change in the whole pedon and has potential for future modeling of carbon cycle over long timescales.

Soil is central in the terrestrial ecosystem, linking and providing feedback responses to the other components, i.e., water, atmosphere, and vegetation. However, the role of soil in landscape evolution is usually not well acknowledged. In modeling landscape evolution, soil is only treated as a residue of weathering that is transported and redistributed along the hillslope. Weathering is considered as a process that produces clays and generates unconsolidated materials available for erosion. While pedology has been debating the form of qualitative factorial models for 75 years; models for soil water, heat, solute, gas and chemical reactions in a profile have matured. As soils are distributed continuously in three dimensions across landscapes, the profile models need to consider lateral fluxes. This review outlines the role of soil in landscape modeling. First, we review the role of soil in the current landscape evolution models. We then review data and models on soil weathering rates and transport processes. We discuss soil profile models that simulate soil formation processes, and combined soil–landscape evolution models. Finally we discuss how the models can be tested and validated in the real world and suggest how both soil scientists and landscape modelers can work together to address the grand challenges in modeling earth surface processes.

Long-term impacts of climate evolution and agricultural management on soil properties and thus soil functioning must be assessed by modelling studies. We focus on modelling of the clay depth distribution as a result of natural soil formation and agricultural management by using the soil genesis model SoilGen2 with loess-derived Luvisols in northern France. An essential first step is the calibration of the processes concerning clay leaching (lessivage) in this model, for which we propose a reproducible method consisting of a screening for a realistic parameter range, a sensitivity analysis and finally a model calibration. To be able to separate the effects of natural soil evolution and more recent agricultural activities, we calibrated the model firstly on natural soils still under forest. Four parameters were the most sensitive and in need of calibration, representing the volume fraction of the soil in contact with macropores, a filter coefficient and physical weathering producing clay-sized particles. These parameters were successfully calibrated, but independent validations at Belgian and Norwegian sites with Albeluvisols and Stagnosols developed in loess and marine clay did not always show an improvement on earlier calibrations. This can be related to several factors, most importantly the dominance of processes other than lessivage in Albeluvisols and uncertain reconstructions of the climate. After calibration, a functional sensitivity analysis was carried out to assess the effect of agricultural management on lessivage. Clear effects on clay loss in the topsoil were found for the length of the crop-cover period and the ploughing depth and to a lesser extent for the intensity of soil mixing and the interception evaporation of the crops. As it is not feasible to reconstruct these factors for the entire period of agriculture in this region (several millennia), further analysis should take the form of scenario studies, but this should be preceded by use of field leaching experiments to calibrate SoilGen2 over short time-scales.

[1] Spatial patterns of soil often do not reflect those of topographic controls. We attempted to identify possible causes of this by comparing observed and simulated soil horizon depths. Observed depths of E, Bt, BC, C1, and C2 horizons in loess-derived soils in Belgium showed a weak to absent relation to terrain attributes in a sloping area. We applied the soil genesis model SoilGen2.16 onto 108 1 × 1 m2 locations in a 1329 ha area to find possible causes. Two scenarios were simulated. Model 1 simulated soil development under undisturbed conditions, taking slope, aspect, and loess thickness as the only sources of variations. Model 2 additionally included a stochastic submodel to generate tree-uprooting events based on the exposure of trees to the wind. Outputs of both models were converted to depths of transitions between horizons, using an algorithm calibrated to horizon depths observed in the field. Model 1 showed strong correlations between terrain attributes and depths for all horizons, although surprisingly, regression kriging was not able to model all variations. Model 2 showed a weak to absent correlation for the upper horizons but still a strong correlation for the deeper horizons BC, C1, and C2. For the upper horizons the spatial variation strongly resembled that of the measurements. This is a strong indication that bioturbation in the course of soil formation due to treefalls influences spatial patterns of horizon depths.

The first results of modeling soil development in marine sediments in S Norway using the model SoilGen are compared to measured properties of two soil chronosequences, on the western and eastern side of Oslofjord, respectively. The aim of this work is to test how well soil development under well-defined environmental conditions can be modeled. Such testing reveals to what degree soil-forming processes are understood, allowing formulation of adequate calculations reflecting these processes. The model predicts particle size distribution reasonably well, although clay depletion in the upper parts of the soils as a result of clay migration is overestimated. The model tends to underestimate contents of organic carbon and CEC in the A horizons: below, modeled CEC matches well with measured CEC. Base saturation is overestimated in the upper 40 cm and underestimated below. Apparently, leaching of bases proceeds less rapidly in reality than is predicted by the model, due to strong soil structure of the B horizons, causing preferential flow and base leaching around the aggregates, whereas bases inside the aggregates are only slightly affected by leaching. Difficulties and possibilities for improvements are identified, some related to model input data and some to the model itself. Input data could be improved by determining the amounts of organic carbon in organic surface horizons and by quantifying effects of bioturbation. A big challenge is the implementation of soil structure formation in the model. Quantitative data on the development of soil structure with time that can be included in a model are required. Amounts, distribution and connectivity of macro pores need to be defined for each stage of soil development, and zones of low and high base leaching need to be distinguished in the model for each time step. The long-term aim of this work is to model soil development with different sets of soil-forming factors, e.g. different climatic conditions in order to reliably predict soil development under different climate scenarios and related sets of soil-forming factors. The results of the first model runs and the identified possible improvements suggest that this aim is generally achievable.

The SoilGen1 model was developed to simulate soil development in calcareous loess at Holocene (15,000 BP–present) temporal extent. We used the LEACHC model as a core and added process formulations to describe the effect of vegetation and soil macrofauna on various soil properties. A limited calibration was done on the calcite solubility constant by comparing decarbonization rates in various leaching climates with values predicted by the metamodel of Egli and Fitze [Egli, M. and Fitze, P. 2001. Quantitative aspects of carbonate leaching of soils with differing ages and climates. Catena 46:35–62.]. Soil profiles of C and pH after 15,000 years compared well with measurements in a Belgian loess profile that was never under agriculture, taking the composition of the C-horizon as uniform parent material. Scenarios with and without agriculture, with varying degrees of bioturbation and for documented Holocene climatic and vegetation evolutions in Belgium and Hungary were simulated and compared. Results show a clear effect of bioturbation on soil development, especially in the continental climate evolution of Hungary. Indicators for the possible occurrence of clay migration and disturbance of lateglacial morphology were calculated and these also show the importance of bioturbation.

Agricultural activities in the Netherlands cause high nitrogen and phosphorous fluxes from soil to ground- and surface water. A model chain (STONE) has been developed to study and predict the magnitude of the resulting ground- and surface water pollution under different environmental conditions. STONE has three main components, namely: 1) the fertiliser distribution model CLEAN; 2) the atmospheric transport and deposition model OPS; and 3) the soil and soil-water quality model ANIMO. The goal of this study was to assess the accuracy of the STONE model output by comparing model predictions with independent observations, while accounting for spatial variation and differences in spatial support. The specific STONE output considered was mineral phosphorous concentration in the top soil. Predictions from STONE are made at the block support and, specifically, for 500 × 500 m grid cells, whereas the observations are on a point support. Therefore, aggregation of the point observations or disaggregation of the STONE model predictions is needed in order to test the accuracy of the STONE model. This study used the aggregation route. Specifically, ordinary and regression block kriging were used to aggregate the point observations to the block support. Overall, there was a good correspondence between the kriged observations and the STONE predictions, with no evidence of bias in the model predictions. However, there were meaningful differences locally, which was partly attributed to the smoothing of the kriging interpolator. Closer inspection revealed that not all of the differences could be attributed to measurement errors in the phosphorous observations and interpolation errors in the kriged phosphorous maps. Thus, it was demonstrated that the STONE output suffers from input and/or model errors. However, quantification of these errors was restricted by the fairly large interpolation uncertainty of the kriged phosphorous observations. Nevertheless, the spatial pattern of error in the STONE predictions could be partly attributed to combinations of landuse and soil type.

Uncertainties in major input variables in water and solute models were quantified and their effects on simulated functional aspects of soil behaviour were studied using basic soil properties from an important soil map unit in the Netherlands. Two sources of uncertainty were studied: spatial variability of basic soil properties, and the use of pedotransfer functions to predict soil hydraulic functions. Depending on the type of process studied, variability of basic soil properties explained between 40 and 99.9% of the total variance. Uncertainty of pedotransfer functions is of importance mainly when physical soil behaviour (e.g. workable days and aeration status) is studied.

Class pedotransfer functions were used to generate average hydraulic characteristics for distinct soil texture classes. Continuous pedotransfer functions were used to generate soil hydraulic characteristics from the actually measured median of the sand particle size, bulk density and percentages clay, silt and organic matter. Both approaches were used to predict the soil physical input data to calculate five different functional aspects of soil behaviour. The functional aspects were: number of workable days, number of days with adequate soil aeration, elapsed time until 10% breakthrough of chloride, amount of cadmium leached after one year and amount of Isoproturon leached after one year. Simulations of water and solute transport were made for 88 profiles which form a statistically representative set of profiles for cover sands in the northeastern part of the Netherlands. The calculated number of workable days did not depend on the type of pedotransfer used. However, the differences between the class and continuous pedotransfer function approach were significant for the other four functional aspects of soil behaviour. For adsorbing cadmium and adsorbing and degradable Isoproturon, differences between the two approaches were statistically significant because they were systematic. However, these differences were so small that they were irrelevant in practice. When to prefer which approach was ambiguous and depended on the functional aspect under consideration. When differences were not significant or irrelevant in practice, the cheaper and easier to use class pedotransfer function approach is preferred over the continuous pedotransfer function approach.

A comité de lecture, avec review; International audience; In this chapter, we discuss hydrology as an input in modeling pedogenesis by comparing the advantages and limitations of different modeling approaches. Water is the main driving force of pedogenesis, as it is responsible for the transfer of solutes and particles and for most of the geochemical and biological reactions occurring in soils. It is thus indirectly responsible for soil texture and structure evolution that influences water transfer and residence time in soils. Two example models that combine geochemistry, hydrology, and plant growth modules are presented: SoilGen and WITCH. These two models simulate the evolution of some soil characteristics on the pedologic time scale reasonably well. However, in their present states of development, these models are not able to simulate any type of soil evolution; therefore, further improvements are needed to implement soil processes and the feedbacks among them. Parametrization and validation, on pedologic time scale, also remain a challenge.

SUMMARYField scale variability in the grain yield of barley in 1989 was investigated in 62 field plots in a Dutch polder area, and compared to soil- and simulation-type characteristics. Total grain mass varied between 3409 and 6019 kg/ha, and grain moisture content between 131 and 14·7%. Soil profile descriptions and soil characteristics were used as basic input data for simulations. Soil water flow was simulated at 119 locations with the LEACHM model, for the purpose of quantifying spatial variability in transpiration deficits in the growing season. Both soil- and simulation-type characteristics were translated from point values to spatial averages for the harvested fields, using kriging. Kriged characteristics were correlated with yields, and used to construct transfer functions. Simulated transpiration deficits during sensitive crop development phases showed negative correlations with grain yield. Transfer functions explained at maximum 68·2% of the variance in the yields.

Thinly stratified sedimentary deposits in a heterogeneous field were investigated to obtain basic physical data for the simulation of water flow. A procedure is described which translates a thinly stratified soil profile into a number of functional layers using functional hydrological properties. A functional layer is defined as a combination of one or more soil horizons and should (i) be recognizable during a soil survey using an auger and (ii) show significantly different functional hydrological properties when compared with another functional layer. This procedure gave three easily recognizable functional layers. Sets of hydrological characteristics of these three functional layers were obtained by physical measurements of the soil and by estimation, using textural data for classification into a standard Dutch series. The performance of several combinations of these sets was tested by comparing simulated and measured soil matric potentials for seven plots during one year. The best simulation results were obtained if measured soil hydraulic characteristics were used for relatively homogeneous functional layers and if the soil hydraulic characteristics were estimated at each location for the most heterogeneous layer.

Soil maps are used to characterize spatial patterns of soil behaviour. A first approach is to characterize each soil map unit (SMU) by the behaviour of a pedogenetically representative profile; a second is to subsample each SMU and select from the sample a characteristic profile by its behaviour. The first approach was tested against the second. Using pedogenetically representative profiles resulted in biased estimations of functional properties related to water and solute transport. Soil profiles fitting to the definition of the SMU (impurities) corresponded to extreme values for four properties, which implies that impurities must be sampled when risk assessments must be made.

Preface. Part I: Keynotes. 1. Soil and water quality at different scales: concepts, challenges, conclusions and recommendations J. Bouma, et al. 2. Relevance of scale dependent approaches for integrating biophysical and socio-economic information and development of agroecological indicators J. Dumanski, et al. 3. Scale issues in agroecological research chains R.J. Wagenet. 4. Obtaining soil and land quality indicators using research chains and geostatistical methods M.R. Hoosbeek, J. Bouma. 5. Some considerations on methods for spatially aggregating and disaggregating soil information A.B. McBratney. Part II: Agroecological and Hydrological Case Studies. 6. Small scale variability in the flow of water and solutes, and implications for lysimeter studies of solute leaching L.P. Simmonds, S. Nortcliff. 7. Solute transport at the pedon and polypedon scales D.E. Radcliffe, et al. 8. Space-time upscaling of plot-based research information: frost tillage H.M. van Es, et al. 9. Mapping and interpreting soil textural layers to assess agri-chemical movement at several scales along the eastern seaboard (USA) T. Steenhuis, et al. 10. Status and trends of soil salinity at different scales: the case for the irrigated cotton growing region of eastern Australia I.O.A. Odeh, et al. 11. Investigating soil and groundwater quality at different scales in a forested catchment: the Waldstein case study G. Lischeid, et al. 12. Slope deposits and water paths in a spring catchment, Frankenwald, Bavaria, Germany K. Arno, et al. 13. Hydromorphic soils, hydrology and water quality: spatialdistribution and functional modelling at different scales P. Curmi, et al. 14. Upscaling a simple erosion model from small areas to a large region D. King, et al. 15. Research on soil fertility decline in tropical environments: integration of spatial scales J.J. Stoorvogel, E.M.A. Smaling. 16. Soil absorbing complex properties of Russian boreal soils and its dependence on the spatial scale of study area J.L. Meshalkina, et al. 17. The influence of nitrate reduction strategies on the temporal development of the nitrate pollution of soil and groundwater throughout Germany: a regionally differentiated case study F. Wendland, et al. 18. Food supply capacity study at global scale J.J.R. Groot, et al. Part III: Methods for Scale Transfer. 19. Upscaling hydraulic conductivity: theory and examples from geohydrological studies M.F.P. Bierkens, J.W.J. van der Gaast. 20. Modelling cadmium accumulation at a regional scale in the Netherlands A. Tiktak, et al. 21. The use of upscaling procedures in the application of soil acidification models at different spatial scales W. de Vries, et al. Part IV: Review Papers. 22. Modelling concepts and their relation to the scale of the problem T.M. Addiscott. 23. Prediction error through modelling concepts and uncertainty from basic data M.J.W. Jansen. 24. Uncertainty analysis in environmental modelling under a change of spatial scale G.B.M. Heuvelink. Part V: Extended Poster Abstracts -- Short Communications. 25. Regression model to predict travel time for chloride leaching through pedons using soil morphological characteristics R. Hatano.

When modelling soil acidification at the European scale, it is inevitable that both model and data have varying degrees of associated uncertainty. The present study attempted to quantify the uncertainty in long-term forecasts of soil solution concentrations of Al{sup 3+} and No{sub 3}{sup {minus}} concentrations resulting from the uncertainty in small resolution European-scale maps and input data, using the Netherlands as a case study. Large-scale forecasts were made with a relatively simple dynamic process-oriented model, SMART2. Model outputs were considered as block median concentrations and the block areal fractions in which concentrations exceeded a critical level. Sources of uncertainty considered included (1) uncertainty in soil and vegetation maps (categorical data), and (2) uncertainty in soil and vegetation-related parameters (continuous data). The uncertainty in model outputs was quantified by an efficient two-step Monte Carlo simulation approach, which takes spatial correlation into account. The uncertainty in the input data at the European scale led to major uncertainties in the predicted Al{sup 3+} concentration. Uncertainties in the areas where the Al{sup 3+} concentration exceeded the maximum allowable concentration were much smaller. The uncertainties in soil-related parameters contributed most to the uncertainty in the Al{sup 3+} concentration, whereas the uncertainty contributed by the soilmore » and vegetation maps was negligible. For the NO{sub 3}{sup {minus}} concentration, however, the uncertainty originated mainly from the soil and vegetation maps. Evaluation of the different error sources is of great practical significance, as it identified which sources need further improvement. The present study shows that the uncertainty contribution of the different error sources depends greatly on the model output considered.« less