INTRODUCTION
Soil degradation has affected more than 1.9 billion hectares of the world's cultivated lands, due to severe-moderate erosion phenomena, caused by surface-water action (Mabit et al., 2014).This, in turn, generates ecological and socio-economic impacts, associated with the reduction of soil productivity, water pollution and the sedimentation phenomena of reservoirs and rivers. Consequently, techniques to determine erosion rates, that offer practicality, certainty and, allow a prospective and retrospective assessment, have a high significance for decision-makers regarding public policies in agriculture and the environment.
The environmental, anthropic and natural radionuclides can be used as tracers because they are deposited in the soil from the atmosphere by precipitation, subsequently suffering the same redistribution due to the different erosion factors to which the soils they adhered to are subjected. The main radionuclides used for this type of study are: 7Be of cosmogenic origin; 137Cs of anthropogenic origin, generated by tests of thermonuclear weapons and accidents in nuclear reactors (Chernobyl, Fukushima and others on a smaller scale); 210Pb of natural origin, generated from the 222Rn decay of the 238U series, part of the 222Rn diffuses through the subsoil to the atmosphere where it decays to 210Pb and eventually precipitates and redistributes in the soils. By measuring the activity of 210Pb and 226Ra parent, it is possible to determine the 210Pexcess that is out of balance with the 226Ra (Zapata, 2002).
Accordingly, using Gamma spectroscopy to measure the activity of 137Cs, 210Pbexc and 7Be, deposited in the soil, even in very low quantities, information to generate distribution patterns, erosion rates, and sedimentation can be determined, at different time scales. Therefore, depending on the radionuclide used, different time spans can be integrated, from a singular event up to a period of three months measuring the activity of 7Be, up to periods ranging from 30 years in the case of 137C (anthropogenic) to 100 years with 210Pbexcess.
Erosion estimates with nuclear techniques, are retrospective. They can be obtained conveniently, making a single sampling at the study sites, without intervening in normal farming operations. Measurements can be scaled from plots to watersheds, without any other consequence than an increase in the number of samples to be analyzed; and in turn, the soils redistribution rates represent the integrated effects of the movement of particulate material, under the uses and management of each unique ecosystem. Hence, nuclear techniques through the use of environmental radionuclides have been widely used for these purposes (Ritchie y McHenry, 1990; Zapata, 2002; Mabit et al., 2014).
The anthropogenic 137Cs radionuclide circulates in the atmosphere following nuclear explosions and accidents that have occurred since 1950. This radionuclide enters the soil through atmospheric precipitation and is strongly retained in the organic-mineral complex of the layers more superficial soils according Mabit et al. (2002) and their vertical migration is very limited (Bunzl et al., 1989). The redistribution balance or total activity of the 137Cs radionuclide per unit area is based on its inventory, measured at a given sampling site and compared to a reference site corresponding to an inventory of accumulation of inputs, taking into account the difference in behavior of cultivated soils (study sites) and non-cultivated soils (reference area). The quantitative estimation of the erosion and sedimentation rates from the measurements of the 137Cs activities requires the use of conversion models that are developed based on the physical processes that influence the interrelations between the magnitude of the reduction or increase in the inventory of 137Cs and land redistribution.
The main objective of this work was to use the 137Cs measurements and the Proportional and Mass Balance-1 conversion models to estimate the net erosion rates in a slope agroecosystem cultivated with horticulture.
METHODS
The study area was located in Waraira Repano National Park, Camino de Los Españoles, next to Caracas Valley, of tectonic origin and formed with Tertiary and Quaternary sediments. It is part of the Venezuelan central coast mountain range, which extends, around four hundred kilometers, parallel to the coast in an east-west direction, from the depression of the Unare to the depression of Yaracuy.
Longitude West | Latitude North | Locations | Reference area and study sites |
---|---|---|---|
66.95169 | 10.543487 | El Fortín de La Cumbre, south slope | Reference Area (RA) |
66.94493 | 10.549852 | Production Unit Hoyo de La Cumbre | Study area, erosion site (ES) |
66.94467 | 10.549769 | Production Unit Hoyo de La Cumbre | Study area, deposition site (DS) |
Table 1 shows the geographic coordinates of the study sites described following:
Reference area (RA) was located in El Fortín de La Cumbre, historically identified as one and the largest of a line of small fortifications that guarded the Camino de Los Españoles; built in 1770, located at 1,428 meters above sea level, with an average annual rainfall of 774 mm according to the record of the meteorological station, serial 5050.
Study sites: The erosion site (ES) is located in Hoyo de La Cumbre, an area under agricultural development since ancient times. It is a horticulture agroecosystem, located 700 meters from El Fortín de La Cumbre (reference area) where slope agriculture has been practiced, with power tiller, sprinkler irrigation and fertilization with NPK formulas 12-24-12 in 2-3 ha extensions. The identification of the deposition site (DS) was carried out in a narrow strip, close to the road bank, in which there is an abrupt change of the slope and the sediments are deposited (Fig. 1). In the DS, preliminary field tests allowed determining an effective soil depth of 80 cm, in contrast to the effective soil depth of 25-30 cm at the erosion site (ES).
Soil sampling: In the RA a systematic sampling of regular distribution (3 m x 5 m) was established, which defined 28 vertices in 7 points and 4 rows. In the study area (ES and DS) a sampling design similar to that of the RA was established, but with spacing of 10 m x 10 m and a total of 26 points. At each central sampling point, 2 replicas were taken at a radio distance of 0.5 m, for a total of 3 sub-samples for each of the sampling depths: 0-20; 20-40 and 40-60 cm, which allowed forming a composite sample for each vertex and study depth. It was only possible to obtain samples at depths of 60-80 cm in the deposit area.
In Figure 1, the changes of the topography in the agroecosystem of horticulture with two inclination planes are showed. They allowed identifying the erosion site with 25% inclination and the deposition site, where sediments from surface runoff was redistributed by the action of hydric erosion. The lower plane, orientated towards the south-east, was localized at 1264.5 m above sea level, whereas the highest plane orientated north-east, was 1279.5 m above sea level, originating a total variation of 15m in the 60m between the altitudinal extremes (Fig. 1 and 2).
The soil cores extracted for each depth: 0-20; 20-40 and 40-60 cm, were dried at room temperature, then homogenized, weighed and sieved, through a 2 mm mesh. The 3 sub-samples of each point were mixed to form a composite sample of each point/depth.
To determine the 137Cs activity, a Canberra gamma spectrometer with hyperpure germanium, HPGe, coaxial detector with relative efficiency of 30% and the Genie 2000 V3.1 program were used. Known masses of the composite samples placed in identified, closed and sealed vessels were positioned in the HPGe detector and measurements were taken in 48 and 72-hour periods.
These measurements were carried out at Nuclear Technology Unit of the Venezuelan Institute of Scientific Research (IVIC).The activities expressed in Bq kg-1 were used to calculate depth of mass relaxation h0, and the activity of 137Cs was determined for the reference area using equation (1):
for undisturbed soils the function may be described by Zhang et al. (1990) and Walling and Quine (1990),
Where:
A’(x) |
amount of 137Cs above depth x, Bq m-2; |
A ref |
137Cs reference inventory, Bq m-2; |
x |
mass depth from soil surface, kg m-2; |
h 0 |
coefficient describing profile shape, kg m-2 |
The Surfer 9.0 program was used to map the local topography and express erosion and deposition estimations in iso-concentrations associated with each vertex of the sampling mesh. To achieve this, precision topography (± 4 mm + 4 ppm) and the point of the geodesic network located at the vertex: Carlota, edo. Miranda, Datum: SIRGAS-REGVEN (10°29'19''.3890 N and 66°50'56''.5679 W) were used.
RESULTS AND DISCUSSION
The reference area 137Cs inventory mean value of 671.69 Bq m-2, was determined. This allowed obtaining the geo-referenced the soil net erosion and deposition rates, using Proportional and Mass Balance-1 Conversion Models of 137Cs measurements to soil net erosion rates (Fig. 3 and 4, respectively). In both figures, the total eroded (negative values) and deposited (positive values) soil special distribution can be appreciated. In both models, the maximum erosion and deposition rates correspond to the areas of higher and lower inclination as well as of higher and lower effective soil depth, respectively. The Proportional and Mass Balance-1 Conversion Models showed maximum erosion rates of 40-45 Mg ha-1 year-1 and 55-65 Mg ha-1 year-1 , respectively. On the contrary, the deposition area, showed accumulation rates of 5-45 Mg ha-1 year-1 and 5-65 Mg ha-1 year-1, for each model respectively. In the proximity to the ramp, where the cultivated area ends, a concave shape is observed, being this characteristic of annual soil gain in which the farmer admits he has higher yields with less fertilization and higher soil humidity (Fig. 1).
The order of magnitude of the maximum erosion rates estimated, agrees with the results reported in the literature using nuclear techniques (Schuller et al., 2003 and Brígido et al., 2006). For south Chilean conditions, for example, where erosion phenomena were studied using 137Cs in an agroecosystem with 19% slope, cultivated and tilled intensely every year, erosion rates between 20 and 70 Mg ha-1 year-1 were estimated (Schuller et al., 2003). Under Cuban conditions, an agroecosystem of undulated plane, with mean inclination between 2 and 5% and maximums of 10%, farmed for more than 30 years, the net erosion rate estimated by Mass Balance-2 model was of 12.2 Mg ha-1 year-1, with 20 cm tillage (Brígido et al., 2006).
In a maize-bean agricultural cultivated area, in volcanic soils of Nicaragua, with annual mean precipitation rates of 1500 mm and slopes of more than 26% , moderate net erosion rate of 28.3 Mg ha-1 year-1 was determined using a metric strip rod (Avilés, 2016). Results in the same order of magnitude were determined in this work. Using Proportional and Mass Balance-1 Conversion Models, the net erosion rates estimated were 19.8 and 29.3 Mg ha-1 year-1, respectively, and a sedimentation velocity, for both models, was of 80%. Silva (2014), registered soil losses between 94 and 135 Mg ha-1 using sedimentation collectors, in a maize agro ecosystem with inclinations of 20 and 30%, respectively.
Comparing estimation methods for net erosion rates, by means of runoff plots opposed to 137Cs with Proportional, Mass Balance-1 and Mass II Balance Conversion Models, Brígido et al. (2006), obtained very similar results between the runoff plots and the Mass Balance-2 Conversion Model. Nonetheless, Proportional and Mass Balance-1 Conversion Models results that could overestimate the net erosion rate.
The advantages and disadvantages of different conversion models have been thoroughly discussed by Walling et al., 2004, and they advise that the Proportional Conversion Model can overestimate soil loss due to loss of 137Cs accumulated on the surface, prior to its assimilation to the profile. In the same way, with the Mass Balance-1 Model, ignoring the 137Cs loss due to recent precipitation events (inducing surface runoff and subsequent erosion), before it is incorporated to the arable layer, as well as the simplification that all the 137Cs deposit occurred in 1963, also induce overestimation in the erosion rates.
Hence, as explained above, by using Proportional and Mass Balance-1 Conversion Models, probably there could be an overestimation of the net erosion rates reported in this work. Nevertheless, the low mean annual precipitation in Hoyo de la Cumbre (774 mm), the predominance of sand fraction, the mean pH (6.5) both, in the reference area and in the study sites, and the management of surface runoff in the study sites, induced to a thought. It is that the causes of the erosion phenomena and the order of magnitude of the estimated net erosion rate are mainly associated to the agricultural management and not to the critical possible overestimation referred.
CONCLUSIONS
For the first time, nuclear 137Cs technique with Proportional and Mass Balance-1 models in the study of the net erosion rate is introduced in Venezuela. The results corroborate moderate net erosion rates associated with intensive horticulture, in slopes with more than 25% inclination and surface runoff management practices. Furthermore, total surface erosion and deposition in the study site were determined, in accordance with the order of magnitude of the net erosion rate and soil redistribution. This allows an integrated view of both phenomena, which are of substantial importance in the formulation, execution, and evaluation of public policies by government and private decision makers on environmental, agricultural, food sovereignty. Thus, it contributes to soil and water assets preservation through the combination of geographic information systems with the benefits of nuclear techniques in the study of soil erosion and redistribution. This technique allows identifying and evaluating the impact of conservation measures of soil and water, in vast extensions of watersheds, where hillside agriculture is widespread, which combined with frequent extreme climate events, generates significant erosion processes and loss of soil fertility with the consequent continuous cycles of intensive fertilizing and pesticide applications, which also contaminate waters.