La stabilité de l'agrégation, un indicateur de la sensibilité des sols au ruissellement et à l'érosion : validation à plusieurs échelles - IRD - Institut de recherche pour le développement Access content directly
Journal Articles Cahiers Agricultures Year : 2001

Topsoil aggregate stability, an indicator of soil susceptibility to runoff and erosion: validation at several scales

La stabilité de l'agrégation, un indicateur de la sensibilité des sols au ruissellement et à l'érosion : validation à plusieurs échelles

Bernard Barthès
  • Function : Author
Eric Roose
  • Function : Author


Several authors have reported relationships between soil susceptibility to erosion and topsoil aggregate stability, which is much easier to assess. However, their estimation of this susceptibility has generally been based on the behavior of 2 mm-sieved samples, which is not always representative of field-scale phenomena. Our aim was to extend the validity of these relationships between topsoil aggregation and erosion, through comparisons of topsoil aggregate stability and field-assessed susceptibility to runoff and erosion. This susceptibility was determined at several levels (Table). lt was first determined through measurements of runoff and soil loss under simulated rainfalls (60 mm/h) on seven microplots (1 m²), in a southern France Regosol mechanically cropped with spring-oat (treatments involved different types of tillage and input). Soil susceptibility to erosion was also determined through three-year measurements of annual runoff and soil loss on runoff plots (700 to 800 m²) in Benin, Cameroon and Mexico (with additional literature data from Syria). The four Benin plots were located on a sandy clay loam Nitosol, subjected to 1200 mm mean annual rainfall and manually cropped with maize (treatments involved different kinds of legume intercropping and input). The five Cameroon plots were located on a loamy sand Ferralsol, subjected to 1300 mm mean annual rainfall, and were under savanna or mechanized and fertilized maize/cotton rotation (treatments involved different cultivation durations, types of tillage and residue management). The five Mexico plots were located on a mechanically cropped loamy Regosol, subjected to 700 mm mean annual rainfall (treatments involved different types of tillage, input and cropping). The six additional Syria plots were located on Regosols and Cambisols, subjected to 300 mm mean annual rainfall, and were mainly under natural rangeland either grazed or protected. Finally, soil susceptibility to erosion was determined on hillside vineyards, growing on various geoogical materials in southern France, through the semi-quantitative assessment of the frequency of erosion features under 600 to 800 mm mean annual rainfall. An erosion index was defined as the sum of weighted (from 1 to 4) frequencies (from 0 to 4) of erosion features observed in the field: sedimentation crusts (weight 1), stones resting on the soif surface (id.), small pedestals (id.), microcliffs (id.), which relate to sheet erosion, and grooves (weight 2), rills (weight 3) and gullies (weight 4), which relate to linear erosion. Soil aggregate stability was determined on 0-10 cm depth samples collected within (runoff plots, hillsides) or close to (microplots) areas where soil susceptibility to erosion had been determined (Figure 1). After air-drying and 2 mm-sieving, 4 g-subsamples were rapidly immersed into deionized water, then wet-sieved through a 0.2 mm sieve with an adapted device. The stable macroaggregate fraction (> 0.2 mm) was calculated as the difference between the fraction > 0.2 mm and the coarse material (> 0.2 mm) content, determined through sieving of the fraction > 0.2 mm into dispersive solution. The fraction < 0.02 mm was determined by sedimentation of the fraction < 0.2 mm (pipette method). The fraction 0.02-0.2 mm was calculated by difference. Four replicates were carried out for each sample. On the southern France Regosol under simulated rainfalls, the runoff depth and soif loss after a 30-min rainfall were significantly correlated with topsoil content in stable macroaggregates (Figure 2). All runoff plots considered (Benin, Cameroon, Mexico, Syria), the mean annual runoff rate and soil loss were significantly correlated with topsoil aggregate stability, especially stable macroaggregate content. Considering slope gradient and climate agressivity (rainfall erosivity) in addition to aggregation led to better relationships (Figure 3). On southern France hillsides, the semi-quantitative erosion index was significantly correlated with topsoil content in stable macroaggregates, especially in the absence of soil conservation practices (Figure 4). These results confirmed that topsoil water-stable macroaggregation, which is known to prevent detachment of easily transportable particles and thereby soil surface clogging and runoff, is a relevant indicator of field-assessed runoff and erosion. A relatively simple and cheap test of aggregate stability therefore allowed a precise assessment of soif susceptibility to runoff and erosion.
Caractériser la sensibilité à l'érosion hydrique des sols en place est souvent coûteux. Plusieurs travaux rapportent une liaison entre la stabilité structurale du sol superficiel, plus simple à mesurer, et la sensibilité du sol à l'érosion ; mais cette sensibilité y est souvent évaluée sur sol tamisé (moins de 2 mm), dont la représentativité est incertaine. L'objectif de notre travail est de valider plus largement la relation agrégation-érosion, en comparant stabilité des agrégats de l'horizon superficiel et sensibilité au ruissellement et à l'érosion, évaluée au champ, à différents niveaux d'investigation. Nos résultats montrent que le taux de macroagrégats (> 0,2 mm) stables à l'immersion dans l'eau d'échantillons secs tamisés à 2 mm est corrélé négativement : au ruissellement et à la perte en terre sous pluies simulées en microparcelles sur Régosol du sud de la France ; aux ruissellements et pertes en terre annuels moyens en parcelles d'érosion au Bénin, Cameroun et Mexique ; à l'intensité des manifestations de l'érosion sur des versants du sud de la France. Un test simple et peu coûteux de stabilité structurale permet ainsi d'évaluer de manière fiable la sensibilité au ruissellement et à l'érosion, pour différents types de sols en place et à différents niveaux d'investigation. (Résumé d'auteur)
Fichier principal
Vignette du fichier
Barthès01_CahAgric_synth agr ero.pdf (4.55 Mo) Télécharger le fichier
Origin Publisher files allowed on an open archive

Dates and versions

ird-04100550 , version 1 (17-05-2023)



Bernard Barthès, Eric Roose. La stabilité de l'agrégation, un indicateur de la sensibilité des sols au ruissellement et à l'érosion : validation à plusieurs échelles. Cahiers Agricultures, 2001, 10, pp.185-193. ⟨ird-04100550⟩


8 View
17 Download


Gmail Mastodon Facebook X LinkedIn More