Scientific Registration n° : 1041
Symposium n° : 31
Presentation : Poster
Soil erodibility assessments with simulated rainfall and
with the USLE nomograph in soils from Uruguay
Estimation de l’érodibilité des sols d’Uruguay sous
pluie simulée et avec le nomographe USLE
VICTORA Carlos, KACEVAS Aarón, FIORI Héctor
Dirección de Suelos y Aguas, Ministerio de Ganadería, Agricultura y Pesca., Av. Garzón
456, 12900 Montevideo, Uruguay
INTRODUCTION AND OBJECTIVES
The maintenance of soil productivity is fundamental for the economy of Uruguay.
Highly productive lands show evidences of past and present erosion. Good estimates of
soil erosion risk could contribute to improve soil management recommendations. Until
now, soil erodibility has been assessed either through subjective judgement based on soil
characteristics or using the USLE's nomograph adapted to local conditions when
numeric erodibility assessments were introduced. Very few studies have measured actual
soil losses due to rain energy.
Measurements of soil losses on the field are difficult and time consuming. Several
methods have been proposed. Standard runoff plots (SRP) and rainfall simulators (RFS)
are the two most common approaches in erosion research.
With SRP, erosion is measured under natural rainfall, but they require time and
their instalation and operation is relatively expensive. RFS are not so precise but their
operation is less costly and they allow multiple replication in short periods of time and in
different sites. The use of specific sprinklers under controlled pressure, allow to
reproduce natural rainfall characteristics (i.e. drop size distribution and drop impact
energy). RFS have been widely used, particularly in developing countries.
The major aims of this paper are: a) to determine relative soil erodibility indexes
(Ks) based on field measurements of soil losses and b) to compare them with those
obtained using the USLE nomograph. Secondarily, relationship of Ks with structural
stability is considered.
In 1932, Duley and Hays used simulated rainfall to measure soil losses. Since
then, there has been a frequent use of RFS as a tool in soil erosion research, including
remarkable designs of the instrument, such as those from Meyer and McCune’s (1958)
with their RFS “rainulator” and Swanson (1965) with his “rotating-boom rainfall
Authors such as Barnett et al (1971) and Dangler and El-Swaify (1977) used RFS
to determine erodibility indexes (K factor) in different soils. A rather wide range of
values for K were reported in these two papers: 0.004 - 0.113 and 0.0 to 0.6
Some studies compared K factors obtained through simulated rainfall and values
obtained using Wischmeier’s nomograph (1971). Young and Mutchler (1977) working
on 13 Minnessotta soils, found that K factors assessed by the nomograph were
overestimated in some soils, while it was underestimated in others. Lago (1981) working
on soils from southern Brazil with a Swanson RFS, reported a K value of 0.022 for a
soil whose computed nomograph-K was 0.156; the K factor measured through SRP (K =
0.0154) was close to the 0.022 RFS-K value. As seen, the relationship between
nomograph-K and RFS-K is not always clear.
In Uruguay, the Wischmeier’s nomograph has not always produced acceptable
values when compared to experts’ opinion. Puentes (1983) estimated the K factor for the
major soils from each of the 99 units of the Reconnaissance Soil Survey Map of Uruguay
using a modification of the nomograph. The author includes a comparison between
estimated K factors and erosion risk from expert opinion, where coincidences and
differences are seen between both estimations
Only one K value was assessed through SRP, with sediment and rainfall data
obtained over a period of 30 months on a Typic Argiudoll (Brunosol Eutrico, Aguas
Blancas, in SE Uruguay). A K value of 0.20 was estimated for this soil (Garcia et al,
Fiori and Martinez (1986) used a low-cost portable RFS (Puentes et al 1979)
designed at the Soil and Water Directorate, based mainly on Swanson's model; they
reported soil loss data (in Mg/ha) measured on field conditions for 3 different soils, but
no K values were computed.
MATERIALS AND METHODS
Characteristics of the Puentes-RFS and procedures. The Puentes-RFS uses 3
sprinklers Vee-Jet H½ U80-100, working at 0.3 kg/cm2 pressure and located at 2.4 m
height, over two duplicate 3.4-x1.0 m contiguous tilled bare soil plots, each one with its
sampling colector system.
Rain intensity is obtained by intermittence, with manual movement. A tin canal
0.15 m wide and 3.4 m long located in between the 2 plots, with a graduated container at
its lowest end, allows to know and regulate rain intensity during the experiment.
A 1,500 L water tank, an electric water pump, an electricity generator and a
wind-protection curtain, complete the set.
The selected rainfall intensity was of 70 mm/h, according to an erosive rain with a
2 yr return period and 30 min duration (Rodriguez, 1984). Each experiment consisted of
3 rainfall events (runs) of 45, 20 and 20 min respectively. In one case (Colonia Brause) a
higher intensity was used (140 mm/h) due to a very high infiltration rate of the soil.
Runoff was measured at regular volumetric intervals, while time was registered
and 1 L samples for sediment load determinations were taken; this was done by
gravimetry. Complementary measurements included: slope, soil moisture at the beginning
of each rain event and structural stability of the surface horizon through wet sieving (the
latter only in 4 soils). The infiltration rate, as the difference between rainfall and runoff,
was computed at the end of the last (3rd) run of each experiment.
It was assumed that: a) rainfall and runoff are simultaneous and b) sediment
concentration in runoff samples is representative of sediment load in runoff occuring in
the time period next before each sampling.
Erodibility estimations. The Wischmeier and Smith (1978) equation was used
to compute rainfall kinetic energy (KE): KE = 11.9 + 8.73 log x, where KE,
Joules/m2/mm and x, rain intensity in mm/h.
The rainfall erosivity index (EI) for each rainfall event was calculated with the
equation: EI = KE.Q.I, where Q is applied rainwater depth (mm) over the soil and I, rain
A “partial” soil erodibility index (Ks, for K-simulator) was assessed for each plot
and for each rainfall event by a linear regression, considering accumulated soil losses and
the rainfall erosivity index (EI). These partial Ks values were corrected for slope gradient
and length (LS). Afterwards, they were averaged in order to obtain the final Ks value for
Soils. Nine representative agricultural soils were selected, all of relevant use for
crop production: 3 Typic Pelluderts and 6 Typic Argiudolls (3 Vertisoles, 1 Argisol and
5 Brunosoles, according to the Uruguayan Soil Classification). Table Nº 1 shows soil
classification according to USDA Soil Taxonomy and to the Uruguayan Soil
Classification, soil Series names and the texture of A horizon (textural families).
Table Nº 1. The studied soils
USDA SOIL URUGUAYAN CLASSIFICATION SOIL TEXTURAL
TAXONOMY (DSA, 1976) SERIE FAMILY
Typic Argiudoll Argisol Subéutrico Ócrico Abrúptico Tomkinson Si
Typic Argiudoll Brunosol ÉutricoTípico Fray Bentos CL
Typic Argiudoll Brunosol Subéutrico Típico Pando SiC
Typic Argiudoll Brunosol Éutrico Háplico Mercedes SCL
Typic Argiudoll Brunosol Éutrico Típico Aguas Corrientes SiC
Typic Pelludert Vertisol Rúptico Lúvico Jesús María L
Typic Pelludert Vertisol Rúptico Lúvico Tala SiC
Typic Argiudoll Brunosol Subéutrico Lúvico Colonia Brause S
Typic Pelludert Vertisol Rúptico Típico Canelones SiC
(*) Si: silty; C: clayey; L: loam; S: sandy.
The Vertisoles, the Argisol and Brunosoles Series Pando and Aguas Corrientes,
were developed from Quaternary silty clay sediments (Libertad Formation). The parent
material for the other 3 Brunosoles were Tertiary loessic sediments for Series Fray
Bentos and Mercedes (Fray Bentos Formation) and sandy sediments for Serie Colonia
Brause (Raigón Formation).
The most significant differences among these soils are in the texture and the
structure of their surface horizons and the presence or not of an argillic subsurface
horizon. The topsoil of Vertisoles, with higher organic matter and clay content are
better structured than Brunosoles and the Argisol, while these soils had a well developed
subsurface clayey horizon (Bt). Soil depths were always more than 80 cm, except in
Serie Mercedes, which was 40 cm.
RESULTS AND DISCUSSION
Table Nº 2 shows soil erodibility indexes assessed through rainfall simulation (Ks)
and using the modified nomograph (K) together with runoff (as a percentage of volume
of rainfall) average sediment load in runoff and structural stability indexes (SSI) when
they were determined.
Table Nº 2. Soil erodibility indexes assessed through rainfall simulation (Ks) and
by USLE nomograph (K)*, runoff, sediment load and structural stability
SOIL ORDER and Ks K RUNOFF LOAD SSI
Serie (%) (AVERAGE) (g/l)
ARGISOL - Tomkinson 0.21 0.73 63 8.8 ----
BRUNOSOL - Fray Bentos 0.17 0.21 40 15.1 2.94
BRUNOSOL - Pando 0.09 0.40 22 7.0 3.21
BRUNOSOL - Mercedes 0.09 0.10 27 9.4 3.03
BRUNOSOL - Aguas Corrientes 0.08 0.36 47 5.1 ----
VERTISOL - Jesús María 0.07 0.21 14 4.3 ----
VERTISOL - Tala 0.03 0.20 43 1.9 5.62
BRUNOSOL - Colonia Brause 0.01 0.21 25 4.1 ----
VERTISOL - Canelones (**) 0.004 0.17 5 1.2 ----
(*) Wischmeier 1971, modified by Puentes 1981
(**) For dry condition
Soil moisture content at 0-15 cm depth for each soil and each run is presented on
Table Nº 3.
Table Nº 3. Soil moisture content (%) at 0-15 cm depth for the 1st, 2nd and 3rd run
with simulated rainfall
SOIL ORDER and SOIL MOISTURE CONTENT % AT 0-15 cm
1st 2nd 3rd
ARGISOL - Tomkinson 20.9 26.1 25.5
BRUNOSOL - Fray Bentos 30.7 31.3 31.7
BRUNOSOL - Pando 24.7 30.3 26.7
BRUNOSOL - Mercedes 27.9 31.0 36.1
BRUNOSOL - Aguas Corrientes 30.6 41.3 41.8
VERTISOL - Jesús María 24.3 21.9 -
VERTISOL - Tala 40.8 43.6 39.9
BRUNOSOL - Colonia Brause - - -
VERTISOL - Canelones 26.9 42.0 48.0
According to the estimated Ks values, we established four soil erodibility groups.
A first group of "high" soil erodibility (Ks > 0.15) comprised by the Argisol
Tomkinson (0.21) and Brunosol Fray Bentos (0.17).
A second group comprises soils with a "moderate" soil erodibility (0.15 > Ks >
0.04); it includes Brunosol Pando (0.09); Brunosol Mercedes (shallow) (0.09); Brunosol
Aguas Corrientes (0.08) and Vertisol Jesus Maria (0.07).
The third group has soils with "low" erodibility (0.04 > Ks > 0.01) and includes
two soils: Vertisol Tala (0.03) and Brunosol Colonia Brause (0.01).
The fourth group, with "very low" soil erodiblity (Ks < 0.01) included Vertisol
Canelones under dry soil condition; this soil had a Ks factor of 0.004.
The elements that might determine the high erodibility index for the Argisol
Tomkinson (Ks = 0.21) are the low permeability of its Bt horizon, that generates a high
percentage of runoff (63%) once the A horizon is saturated, its high silt content and the
poor structure of the surface horizon.
On the other extreme of the erodibility range, the very low erodibility of Vertisol
Canelones, is probably due to a highly stable structure and a low runoff volume (5%).
The latter was a result of its high infiltration rate (42 mm/h). In this case, this was
favoured by the (seasonal) low initial topsoil and subsoil moisture content.
The low Ks found for Vertisol Tala in spite of its high runoff, was probably due
mainly to a good structure of the topsoil, while for Brunosol Colonia Brause its low Ks
was attributed to its very high infiltration rate (105 mm/h) as a result of the sandy texture
of its A horizon; these two characteristics significantly decreased runoff, and so, the Ks
value (rain intensity applied was increased up to 140 mm/h, so to generate runoff).
In all cases Ks were lower than K. Furthermore, assessed Ks does not show a
correspondence with K values. This lack of correspondence is probably due to the
complex interaction of many soil properties and factors such as those related to natural
and simulated rainfall characteristics that affect soil losses under field conditions.
Examples of this discrepancy are Brunosol Pando and Vertisol Jesus Maria.
Since Brunosol Pando has a higher silt content, lower organic matter content,
lower infiltration rate, and a more intensive agricultural use (crops vs. pasture) than
Vertisol Jesus Maria, higher soil losses and a markedly higher K factor were expected in
the former. However, experimental data showed that soil losses and Ks were similar for
both soils (0.09 and 0.07 respectively). This fact could be explained by the formation of a
stronger soil crust and a flat, non rugose ("ironed") surface on Brunosol Pando, in a
shorter time than what it takes in Vertisol Jesus Maria. That crust and micro-topography,
might confer surficial resistance to the soil against raindrop impact and protection against
accelerated detachment and transport of soil particles. Then, the protective effect of the
water runoff layer prevents further detachment and transport of remaining soil
aggregates and particles on that flat, highly resistant soil surface. Differences in the
thickness of the crust and in the remaining micro-topography were easily seen at the end
of each experiment.
Probably these interactions and processes are not thoroughly accounted for by
the USLE's nomograph for local conditions such as those from several cases in Uruguay.
Initial soil moisture content in each rain application did not explain differences in
between soils except for one case, the Vertisol Canelones, where the dry season and
condition of the soil probably generated a high infiltration rate and low runoff. It was the
soil with the greatest difference in water content between the second (24%) and the third
run (40%) (Table Nº 3). This soil Order would justify to be considered with two
erodibility indexes according to moisture content (summer, dry; winter, wet) due to very
important differences in their soil losses in each case, as was proposed by Puentes
A good correspondence between soils SSI and sediment load in runoff was
observed in the four cases where this index was determined. Correspondence of SSI with
the Ks values was not so clear, although they showed an inverse tendency (higher SSI,
lower Ks) as shown in Table Nº 2.
Soil erodiblity indexes (Ks) here determined by rainfall simulation show
significant differences among soils and possible cause-effect reationship between this
attribute and certain soil characteristics.
A good correspondence was observed between structural stability indexes and
sediment load in runoff water, together with an inverse tendency of those indexes with
the Ks values.This suggests that SSI could be a less costly and reliable way to estimate K
factor. More research on this relationship, including soils with contrasting topsoils’
properties, could validate this observation.
No clear relationship was found between estimated Ks and K computed by means
of the USLE's nomograph. We attributed this lack of correspondence to site specific
conditions and to the known complexity of the soil erosion process. Results indicates
that one must be careful when using the latter method to estimate K factors, at least for
soils in Uruguay, and consequently, to predict soil erosion losses. At the same time, data
here presented could contribute to improve the results of the use of the USLE in this
country and thus, to improve soil conservation practices and agricultural production
The authors are grateful to Dr. Juan Burgueño at Facultad de Agronomía,
Montevideo-Uruguay, for his help in statistical analysis, and to Dr. Ruben Puentes at
Rockefeller Foundation, México-DF and Dr. Ariel Szogi at the Agricultural Research
Service-USDA, Florence-SC, for their assistance in reviewing the manuscript. They also
express their thanks to the staff of the Soil and Water Directorate for their help in the
field experiments, laboratory analysis and typing .
This study was part of a research program carried out within the framework of a
technical cooperation agreement between the Dirección de Suelos y Aguas of the
Dirección General de Recursos Naturales Renovables - Ministerio de Ganadería,
Agricultura y Pesca, and the Programa de Manejo de Recursos Naturales y Desarrollo
del Riego (PRENADER-BIRF).
1. Barnett, A.P., J.R. Carreker, F. Abruna and A.E. Dooley (1971) - Erodibility of
selected tropical soils. Trans. ASAE 1:496 - 499.
2. Dangler, E.W. and S.A. El-Swaify (1976) - Erosion of selected Hawaii soils by
simulated rainfall. Soil Science Society of America Proceedings 40(5): 769 - 773.
3. Duley, F.L. and O.E. Hays (1932) - The effect of the degree of slope on runoff and
soil erosion. Jour. Agr. Res. 45:349 - 360.
4. Fiori, H.H. y S. Martínez (1986) - Pérdidas por erosión y medidas de infiltración
con lluvia simulada en tres suelos agrícolas del Uruguay. Tesis, Facultad de
Agronomía, Montevideo, Uruguay.
5. García, F., G. Cardellino, J. Pannone y A. Szogi (1984) - Metodología empleada en
parcelas de escurrimiento y avance de resultados obtenidos en la Unidad
Experimental de Aguas Blancas. IICA-Ministerio de Agricultura y Pesca.
Montevideo, Uruguay (Mimeografiado).
6. Lago, J.C. (1981) - Erodabilidade de um solo podzólico, vermelho amarelo
eutrófico, pelos métodos da chuva natural, simulador de chuvas e nomograma, no
Sertão de Pernambuco. MSc thesis, Universidade Federal de Pernambuco, Brasil.
7. Meyer, L.D. and D.L. Mc Cune (1958) - Rainfall simulator for runoff plots. Agr.
Eng. 39:644 - 648.
8. Puentes, R., C. Víctora, A. Szogi, D. Sciarra, G. Acosta, H. Fiori y S. Martínez
(1979) - Diseño y calibración de un simulador de lluvia de bajo costo. 14 pp
(mimeografiado) Dirección de Suelos y Aguas - MGAP.
9. Puentes, R. (1981) - A framework for the use of the Universal Soil Loss Equation in
Uruguay. M.Sc. Thesis, Texas A&M University. College Station, Texas.
10. Puentes, R. (1983) - Una metodología para evaluar la capacidad de uso de las
tierras. Dirección de Suelos, Programa MAP-IICA-INC, Montevideo-Uruguay.
11. Rodríguez, A. (1984) - Fórmulas Lluvia-Duración-Retorno-Riesgo en las ocho
subregiones del Uruguay. Agua en la Agricultura Nº 2 (2ª Edición). MGAP-
DGRNR-DSA. pp 19-24.
12. Swanson, N.P. (1965) - Rotating boom rainfall simulator. Trans. ASAE 8(1): 71-
13. Wischmeier, W.H., C.B. Johnson and B.V. Cross (1971) - A soil erodibility
nomograph for farmland and construction sites. J. Soil and Water Cons. 26:189 -
14. Wischmeier, W.H. and D.D. Smith (1978) - Predicting rainfall erosion losses: a
guide to conservation planning. Agricultural Handbook Nº 537, USDA,
15. Young, R.A. and C.K. Mutchler (1977) - Erodibility of some Minnesota soils. J.
Soil and Water Cons. 32:180 - 182.
Keywords: erodibility, rainfall, rainfall simulator, Vertisols, Argiudolls, USLE, soil
erosion, soils, nomograph, Uruguay
Mots clés : érodibilité, pluie, simulateur de pluie, Vertisols, Argiudolls, USLE, érosion
des sols, nomographe, Uruguay