Spatial analysis of the soil erosion in the hydrographic basin of Formoso River: subsidy to politics of land use and environmental preservation in a region of Brazilian Cerrado

The erosion of the soil is one of the most worrisome environmental problems of the planet, which causes negative impacts as fertility loss of lands and water bodies pollution. The objective of this paper was to analyze the soil erosion in the hydrographic basin of Formoso River, located in the state of Goiás, Brazilian Cerrado. Most part of the hydrographic basin area belongs to the Emas National Park Conservation Unit. To estimate the soil loss from surface erosion, remote sensing products were used, geoprocessing tools and the Universal Soil Loss Equation (USLE). The annual erosion values in the basin varied from 0 t ha -1 year -1 to 8315 t ha -1 year -1 , with 85.88% of the total area for soil losses classified in the slight category. The highest soil losses are associated with high results for R, K and LS factors in agricultural use, which represents 29.12% of the total area, with losses of 97.78% of soil from real erosion, in soils classified dystrophic Red Latosol, medium texture and high topographic values areas. The agricultural areas and the analysis of the real erosion reassures the necessity of proper planning, considering correct management techniques, in order to it won’t compromise the water resource. This study could subsidize the planning and management of the referred hydrographic basin and this important preservation area of the Cerrado.


Introduction
The erosion is a superficial process and it may be defined as the detachment and transportation of the soil particles, being the main cause of deterioration and accelerated impoverishment of the lands (Santos et al., 2019).The destructive effect of erosion influences human life in several ways, driving it to fertility loss, the food security, truncation of the soil, instability of slopes, causing irreversible impacts in the little renewable soil resource and compromising water resources (Abdulkareem et al., 2019;Mohammed et al., 2020;Gebrehiwot, 2022).
Soil loss by anthropic agents occurs especially due to improper agricultural practices, urbanization, deforestation, forest fires, infrastructure deployment and excessive pasture (Endalew and Biru et al. 2022).
Brazil, in spite of being a young country, it already shows signals of evident severity of the problem of fertility decrease in some regions (Bertoni and Lombardi Neto, 2014;Gomes et al., 2019;Medeiros et al., 2022), due to the intense and improper use of the soils, compromising biomes, and the quality of life.Among the biomes affected by anthropic actions, Cerrado stands out, which, since the 1970s, passed through an intensification of soil use.The state politicies did region one of the main focuses of the farming development and expansion programs in Brazilian inland.Accordingly the farming development strategies implemented in Brazilian Cerrado, it's possible to mention the creation of Cerrado Agricultural Research Center's (CPAC) from Brazilian Agricultural Research Corporation (EMBRAPA) and the launching of Japanese-Brazilian Program for the Development of the Cerrado (PRODECER), a partnership with Japanese Agency of International Cooperation, Program for the Development of the Cerrado (POLOCENTRO) (Pessôa, 1988;Rocha et al., 2014;Martins et al., 2016).
Thus, Brazil and Japan signed, through the ODA, technical cooperation agreements, whose four main modalities were represented by a) Project-Type Technical Cooperation; b) Studies for Development; c) Investment and Financing for Development; and d) Financial Cooperation, which financed projects aimed at the agricultural sector, in the form of loans in Japanese currency, related, in most cases, to the implementation of infrastructure for the agricultural production system in the Cerrados (Inocêncio, 2016).The program was implemented in three phases: PRODECER I pilot, from 1979PRODECER I pilot, from to 1983;;PRODECER II Pilot andExpansion PRODECER, from 1985 to 1993;and PRODECER III Pilot, which started in 1995and ended in 2001. PRODECER I pilot project (1979-1983) was initially implemented in four areas: Coromandel, Mundo Novo, Iraí de Minas and Entre Ribeiros, in western Minas Gerais, incorporating an area of 60 thousand hectares, with 92 settled families, benefiting medium-sized farmers (between 250 and 500 hectares).In the second phase (PRODECER II Pilot (1985-1990) and PRODECER Expansion (1985Expansion ( -1993)), the program was subdivided into two stages.The first stage, considered as pilot areas, in Cerrado spaces with different natural conditions, in the states of Mato Grosso and Bahia, implemented in four projects with actions in areas considered pilot and in Cerrado spaces, with different natural conditions in the states of Mato Grosso and Bahia.The PRODECER III Pilot (1995Pilot ( -2001) ) represented the continuation of a work aimed at incorporating new areas of Cerrado in Brazil to modern agricultural activity outside the axis and proximity of the Center-West region of the country.In this stage, two spaces with Cerrado were incorporated, in two projects of 40,000 ha, with 40 producers settled in each one, in the municipalities of Balsas (south of the State of Maranhão) and Pedro Afonso (State of Tocantins) (Santos, 2016).
Allied with PRODECER, the main transforming agent in the rural environment in the region of Cerrado was technology, represented by modern management techniques and production control, which enable the high productivity indexes with the obtention of more than one harvest a year.In addition, the opening and paving of highways, and the investment from the Federal Government in occupation programs and farming production of Midwest and North of Brazil had placed the state of Goiás in the economic map of Brazil (Freitas, 2013;Martins et al., 2016;Mata et al., 2017).The modernization of the Brazilian agrarian space also generated negative impacts on the national land structure by repelling small farmers, without capital to incorporate new technologies, thus accentuating land concentration and social inequality in the countryside (Cabral et al., 2023).
The policies of land use and occupation associated with the technological advancements have transformed the Cerrado into a Brazilian breadbasket, and they estimate that over 20 million hectares (corresponds to 9,8% of the total area of the Brazilian Cerrado) are developed agricultural activities, while 56 million hectares (corresponds to 27,5% of the total area of the Brazilian Cerrado) are occupied by pasture (Latrubesse et al., 2019;MAPBIOMAS, 2017).This expansion, although had increased the offer of food, it also represents an extreme fragmentation of Cerrado (Hunke et al., 2015;Latrubesse et al., 2019;Parente et al., 2021;Resende et al., 2021;Santos and Naval, 2020).In Goiás, for example, about 69% of the hydrographic basins bigger than 500 km² have less than 50% of vegetation remaining (Latrubesse et al., 2019), and even 46% of those basins do not fulfill the requirements of the Brazilian Forest Law (Law 12,651, of May 25, 2012), which establishes norms for the protection of native vegetation in areas of permanent preservation, legal reserve, restricted use, forest exploitation and related matters (BRASIL, 2012).In addition, the recent changes in the Forest Code, and also the amnesty for environmental crimes harmed the legal situation of 7.6 million hectares of riparian forests of Cerrado (Latrubesse et al., 2019).
Therefore, studies concerning soil erosion are important to subsidize actions aiming at environmental preservation.According to Federal Law no.9,433/1997, the hydrographic basin is the territorial unit for the implementation of the National Policy of Water Resources and the performance of the National System of Management of Water Resources (BRASIL, 1997).In addition, the hydrographic basins are better units for studies and planning of soil loss, once they are used in regional scales which give great spatialization of the phenomena that take on site (Castro et al., 2022).
As the direct measurements of erosion are expensive and take too much time, the elaboration of soil loss prediction models based on remotely acquired data has received great attention from soil scientists (Batista et al., 2017;Alves et al., 2021).One of the most used equations nowadays, mainly due to its facility of implementation and by its diffusion in the world, is the Universal Soil Loss Equation (USLE), developed by Wischmeier and Smitth (1978).Research like the ones accomplished by Ali andHagos (2016), Fernandez et al. (2018), Pham et al. (2018) have demonstrated how it has been used in different scales and some countries.With the progress of geotechnologies, the elaboration of spatial simulations has become viable, which allows taking into consideration the spatial variations of the conditioning factors of the erosive process (Coutinho et al., 2018).
The geotechnologies can be defined as a set of scientific methods and technologies to collect, process, analyze, and offer information with geographic reference (Barros Júnior et al., 2018;Oliveira et al., 2023), being composed by of the Remote Sensing (RS) system, Geographic Information System (SIG), Global Navigation Satellite System (GNSS), geoprocessing, digital cartography, and the like (Dixon et al. 2015).
Among the most used ones, we can emphasize RS and SIGs.With the advancement of RS and SIG, it's possible to analyze and interpret, for example, the USLE factors, which are erosivity (factor R), erodibility (factor K), relief dissection (factor LS), land use and cover (factor C) and conservationist practices of the soil (factor P).From the obtention of these factors, it's possible to estimate soil loss from surface erosion (Alves et al., 2022).
Thus, the objective of this study was to understand the relations among rainfall, relief aspects, soils, land use and cover, and soil loss from surface erosion in the hydrographic basin of Formoso River, using the USLE model adapted by Lombardi Neto and Moldenhauer (1992) for Brazilian environmental conditions from the model proposed by Wischmeier and Smith (1978) (Bertoni and Lormbardi Neto, 2014), being the process of modeling accomplished through geotechnologies.The hydrographic basin of Formoso River was chosen for this study due to its economic and environmental importance, once it comprises the greatest part of Emas National Park (PNE), Moreover, nother highly relevant factor is due to be strategic source of water supplying, to Chapadão do Céu city of, which is estimated in 10,797 people (IBGE, 2021).
The PNE is one of the few Preservation Units that contain different Cerrado phytophysiognomies in Goiás state, Brazil, as dirty fields, clean fields, vereda, and riparian forests, besides the characteristic landscape, it's an important to observe typical animals from Cerrado (ICMBio, 2019), tourism, and scientific researches.However, even with its notable importance, has been going through concerns, such as the elimination of the surrounding vegetation, due to the intensive agricultural expansion, compromising the biota, and the ecosystemic balance, once it disables the formation of biodiversity corridors (Soares et al., 2016), among others environmental problems.Thus, technoscientific data based on studies concerning soil loss are necessary to subsidize the implementation of effective and efficient public policies, aiming at the proper management of the study area and the preservation of the PNE.

Materials and mthod
All the methodological procedures were accomplished through ArcGIS Advanced 10.8.1® software, licensed by code #647261 (ESRI, 2020).The maps were elaborated using the Universal Transverse Mercator (UTM) projection, and the Geographic Coordinates System, Datum Geocentric Reference System for the Americas (SIRGAS) 2000, 22S Zone South (S).

Study area
The hydrographic basin of Formoso River is located in the municipalities of Mineiros and Chapadão do Céu, in the Southwest microregion of Goiás state -Brazil.With an area of about 1,236km², 67.3% (831.73 km²) belonging to the municipality of Mineiros (GO) and 32.7% (404.27km²) to the municipality of Chapadão do Céu (GO).The two municipalities present a population estimated at 68,154 and 10,167 inhabitants, respectively, which together result in a population estimated at 78,321 inhabitants (IBGE, 2020).As mentioned before, it comprehends the greatest part of the PNE (Figure 1).The study area is located in the Cerrado biome, which regions vary from one unique geomorphological compartment to others quite heterogeneous with a high degree of dissection, being composed of plateaus and depressions.The plateaus domain the Center-South region of the biome with average elevations between 528 m and 1045 m, while the depressions domain the Center-North region with average elevations between 85 m to 415 m (Sano et al., 2019).The Southwest of Goiás is characterized by the continuity of the geological and structural conditions of the Sedimentary Basin of Paraná, formed by plateaus and chapadões (tablelands) in levels with smooth valleys which intermediate flat backgrounds or slightly carved in V, when they show maturity.The geology of the hydrographic basin of Formoso River is composed of lithologies dated from the Neogene and Cretaceous, being formed by volcanic rocks (São Bento Group -Serra Geral Formation) and sedimentary ones (Unit with undifferentiated detrital coverage, Bauru Group -Peixe River Valley Formation, Alluvial deposits, and cachoeirinha Formation) (SIEG, 2021).
The vegetation of Cerrado presents eleven phytophysiognomies: gallery forests, riparian forest, dry forest, cerradão, stricto sensu Cerrado, dirty field, clean field, palm trees, Cerrado park, veredas, and rupestrian fields.In the hydrographic basin of Formoso River, the following phytophysiognomies were observed: Cerrado stricto sensu, Cerradão, Riparian Forest/Gallery, and Rural formation.These phytophysiognomies can be arranged in three main formations: forestry, savannah, and rural (Santos et al., 2020).The state of Goiás is dominated by formations that characterize the Savannah, due to the presence of shrubs, and physiognomies that contemplate other types of vegetation, as forest areas, that are distributed in spots, islands, or corridors (Oliveira, 2014).
In the region of the study area, although more than half the soil is chemically limited by its low reserve of nutrients, and high levels of aluminum, because of its need to be corrected for the development of agricultural activities, it was verified that 48% of the lands are mechanized with the domain of several types of Latosols, mainly Red Latosol with clayey texture, with flat and smooth wavy relief.About 1.5% consist of improper areas for any kind of cultivation due to gravel and rocky aspects of Litholic Neosoil and Cambisol types (Oliveira et al., 2013).
According to Cardoso et al. (2014), using the method proposed by Köppen-Geiger in 1900 andadapted by Setzer, (1966), with data generated by the study of two climatic variables (air temperature and rainfall), the climatic classification of the study region (hydrographic basin of Formoso River) is included on AW type, tropical climate and with the dry season in the winter.The climate performs an important influence concerning the shape of the relief, being one of the most relevant elements in its modification and transformation (Cardoso et al., 2014).Among the climatic elements, rainfall stands out.In Figures 1 and 2 the rainfall of the study area is presented.For the results obtained from the results used (ANA, 2019), the information about the acquisition and the organization are explained further on, in the results were 2.2 Databases and 2.4 Rainfall and Factor R.

Universal Soil Loss Equation
To estimate the soil loss from surface erosion, the USLE adapted by Lombardi Neto and Moldenhauer, (1992) was used for the environmental conditions of Brazil, from the model by Wischmeier and Smith (1978).The adapted model enables estimation of the annual average soil loss in varied conditions of climate, soil, relief, land use and cover, and conservationist practices, so to indicate areas that present fewer or greater soil losses (Alves et al., 2022).According to Equation 1: At which: A is the soil loss from surface erosion (t ha -1 year -1 ); R is the erosivity factor of the rainfall (MJ mm ha -1 h -1 year -1 ); K is the erodibility factor of the soil (t h MJ -1 mm -1 ); LS is the topographic factor, considering the length of ramp (L), and the declivity (S) (dimensionless); C, the factor of use and cover of the soil (dimensionless); and P, the factor of conservationist practices of the soil (dimensionless).
The R, K, and LS factors are associated with natural aspects of the study area, rainfall, topography and soil, respectively.The C and P factors are distinguished from the other ones for being related to anthropic actions (Alves et al., 2022), besides the cover related to the native vegetation.

Rainfall and Factor R
The rainfall data (monthly average and annual average) was complied from HidroWeb website (ANA, 2019) of 6 stations (Table 1, and Figure 2) in the surroundings and inside the hydrographic basin study (Table 1).Database were organized in an electronic spreadsheet and treated (filling gaps through the average).The erosivity was calculated for each station, according to the equation by Wischmeier and Smith (1978)) adapted by the Brazilian environmental conditions by Lombardi and Moldenhauer (1977) apud Bertoni and Lormbardi Neto (2014): At which: R is the erosivity factor of the rainfall (MJ mm ha -1 h -1 year -1 ); ri, the monthly average rainfall (mm); and P, the annual average rainfall (mm).
The Erosivity map was generated through the interpolation of data obtained from the equation (2).The interpolation is a method capable of getting specific values from two points or more.In this study, the interpolation method was applied through the spline tool of ArcGis.The results were compared with the classification proposed by Carvalho (2008) (Table 2).Source: Carvalho (2008).
Factor K The soil map was elaborated from the Soils Map from the Master Plan of the Basin of Paranaíba River (SIEG, 2017), updated in SIG environment according to Santos et al. (2018).To obtain the erodibility values reported in the literature have been used, according to Corrêa et al. (2015); Demarchi and Zimback (2014); Lima et al., (2016).The erodibility values were compared to Mannigel et al. (2002) classification, according to Table 3. Factor LS To estimate the LS factor (direction data and flow accumulation), the basin delimitation was used to extract the SRTM from the management unit.Moreover, was use the flow direction tools, flow accumulation, slope and reclassify.The SRTM model comes from a project that collected topographic data at a global level, enabling the generation of digital elevation models with a high degree of detail (Melo et al., 2020;Neves et al., 2021).Thus, to obtain this factor we've used the SRTM model, and the method proposed by Pelton et al. (2012), using Equation 3 in the ArcGIS raster calculator.LS = Power(af * rc/22.1,0.4)* Power (sin (d * 0.01745)/0.09,1.4)* 1.4 (3) At which: LS is the topographic factor; af, the flux accumulation; rc, the resolution of the cell; and d, the declivity.

Factor CP
The factors C and P of the USLE were integrated and formed a unique cartographic product (CP), considering the factor P equal 1, as observed in studies by Alves et al. (2022); Barbosa et al. (2015); Lima et al. (2018).Thus, the spatialization of the results for the CP factor is directly linked to the land use and cover of the hydrographic basin of Formoso River, with the indexes defined as observed in Barbosa et al. (2015); Bertoni and Lombardi Neto (2014); Durães and Mello (2016); Helfer et al. (2003); Lima et al. (2015); Silva (2004).
From images of the MSI sensor, the land use and cover map was elaborated, using the Interactive Supervised Classification classificatory tool, algorithm of interactive and supervision classification, and manual corrections to better represent the study area.We've used the bands B2 (blue), B3 (green), and B4 (red), and B8 (nearinfrared).In the land use and cover classification process, eight classes were used, such classes and their respective number of samples are: (i) cultivated landmainly soy and corn interim harvest (1,181 samples); (ii) cerradão (1,174 samples); (iii) stricto sensu cerrado (261 samples); (iv) riparian forest and gallery forest (486 samples); (v) rural buildings (374 samples); (vi) rural formation (589 samples); (vii) pasture (99 samples); and (vii) water (374 samples).It was not decided to use the eleven classes of Cerrado phyto physiognomies for the elaboration of the land use and land cover map.The definition of the legend was made from the previous knowledge of the study area, along with the analysis of images from MSI sensor, used in the classification process, and with the help of high-resolution images enabled and visualized through Google Earth Pro (Google, 2019).After the classification process, we've accomplished topographic corrections through the Topology tool from ArcGIS.
For the map validation, a confusion matrix was made and the Kappa Index was calculated.This index is a statistical test proposed by Jacob Cohen in 1960 (Cohen, 1960), which is used to evaluate the quality of the classification of land use and cover.It takes into consideration all the confusion matrix in its calculus, including the elements out of the main diagonal or descendent diagonal.The Kappa Index was estimated through Equation 4, a simpler method to obtain this index observed in Figueiredo and Vieira (2007).
At which: K is an estimative of the Kappa coefficient; n, the total number of samples; c, the total number of classes; xii, the value of line i and column i, in other words, the diagonal value of the confusion matrix, in a descendent way; xi+, the sum of line i; and x+i, the sum of column i from the confusion matrix.

Potential erosion and real erosion
The results for the potential erosion were obtained through the combination of the factors R, K, and LS, through map algebra, with a raster calculator tool.The results were classified and interpreted according to the proposal by Valério Filho (1994) (Table 4).The real erosion was calculated considering the integration between the potential erosion and the CP factor.To better understand and analyze the results of real erosion, the values found were classified and interpreted according to the proposal by Beskow et al. (2009) (Table 5).Extremely high Source: Beskow et al. (2009).

Potentialities and soil losses
In Table 6 the results of the erosivity obtained for each rainfall station are presented, varying from 8,124.14 MJ mm ha -1 h -1 year -1 to 8,927.78MJ mm ha -1 h -1 year -1 .It's important to emphasize that at the station which is located at Emas National Park (PNE) (Formoso Farmcode 1852001), the erosivity found was 8,483.11MJ mm ha -1 h -1 year -1 .This value is about 5% less than the greater value found in the region (station at São Bernardo Farm, located in the North surroundings of the studied basin).
In Figure 5A are presented the geospatialized results of the erosivity, with the variation from 8,432.90MJ mm ha -1 h -1 year -1 to 8,575.80MJ mm ha -1 h -1 year -1 , being classified according to Carvalho (2008), as strong erosivity.It's possible to observe a greater erosivity in the North region of the highest part of the hydrographic basin, due to a greater rainfall index in this region.In Table 7 the soil classes are presented with their respective textures, areas, values for erodibility, and the consulted sources.In Figure 5B and Figure 5C the results of the soil and erodibility classes present in the studied hydrographic basin are presented, respectively.The area of the basin presents four soils classification, where there's the predominance of Dystrophic Red Latosol with clayey texture or very clayey (DRL) (74.58%), followed, in decending order, by Dystrophic Haplic Gleysol (DHG) (14.60%),Dystrophic Red Latosol with medium texture (DRL2) (10.59%), and Quartzarenic Neosol (QN) (0.23%).According to the classification by Manningel et al. (2002), the erodibility results vary from extremely high to low.The greatest erodibility value is associated with soil Quartzarenic Neosol (0.084 t h MJ -1 mm -1 ), classified extremely high erosivity, present in the part of the hydrographic basin (region North). he second greatest erodibility value was associated with soil dystrophic Red Latosol with medium texture (0.046 t h MJ -1 mm -1 ), observed in the lower part of the hydrographic basin.The lower erodibility values is associated with soils dystrophic Haplic Gleysol (0.020 t h MJ -1 mm -1 ), present mainly in the valleys and dystrophic Red Latosol with clayey texture or very clayey (0.0131 t h MJ -1 mm -1 ), observed in the high and medium part of the hydrographic basin Formoso River (Table 7, Figure 5B Figure 5C).
In Figure 5D the results of the LS factor are presented, which vary from 0 to 1,242.50, with an average of 0.045.We've observed that greater values are associated with nearby areas from the drainage network.The factors R, K, and LS were integrated, generating the cartographic product about potential water erosion, from 0 t ha -1 year -1 to 322,797.53 t ha -1 year -1 .According to the classification proposed by Valério Filho (1994), 95.5% of the total area is classified in the categories of Weak erosion (0 -400 t ha -1 year -1 ), and Moderate (400 -600 t ha -1 year -1 ), and 4.5% in the Medium (600 -1600 t ha -1 year -1 ) to Very strong (> 2400 t ha -1 year -1 ).The greatest results of the potentiality of soil loss were found near the drainage of the hydrographic basin, in the lower basin, associated with high values of the topographic factor (LS), and the erodibility factor value of the dystrophic Red Latosol with medium texture (Figure 5C, Figure 5D, Table 8, and Figure 6).
From the total area of the hydrographic basin, about 70.10% is covered by native vegetation (correspond: Cerrado stricto sensu, Cerradão, riparian forest/gallery, and rural formation), 29.12% is destined to cultivated land (mainly soy and corn), and 0.11% to pasture (Table 10).The values related to the CP factor vary from 0 to 0.2, with the lower values related to categories that expose less soils to erosive processes (e. g., water and rural buildings), and the greater ones, to classes that let the soils exposed to soil erosion (e. g., cultivated land and pasture) (Figure 7, Table11 and Figure 8).The results of the integration of the natural potentialities to the erosive processes (potential erosion), and CP factor, originating the real erosion, are presented in Table 12, and Figure 9.The values varied from 0 to t ha -1 year -1 to 8.315 t ha -1 year -1 , with 88.99% of the total area in the categories Slight (0 -2.5 t ha -1 year -1 ) to Moderate (5 -10 t ha -1 year -1 ), and 11.01% in the categories Moderate -high (10 -15 t ha -1 year -1 ) to Extremely high (> 100 t ha -1 year -1 ).In Table 13 the results of real soil loss by category of land use and cover, and the annual total in t ha -1 year -1 are presented.For the year of 2019 observed an estimate of soil loss in the hydrographic basin of Formoso River of 14,242,639.52 t ha -1 year -1 , emphasizing the cultivated land, that even representing 29.12% of the total area, has lost 97.78% of all the annual soil loss of the study area.

Discussion
About the erosivity of the rain, in a study also made in the state of Goiás by Lima et al. (2018) in the Metropolitan region of Goiânia (capital of Goiás state), there were similar results to the ones obtained for the basin of Formoso River, with annual average values varying from 8,157.22 MJ mm ha -1 h -1 year -1 to 8,826.89MJ mm ha -1 h -1 year -1 , classified as strong erosivity, according to the classification proposed by Carvalho (2008).
Evaluating the intensity and distribution in space-time of the erosivity of the rain in the state of Goiás and Distrito Federal, Anjos et al. (2020), a variation of 4,514±2.5 to 11,215±1.8MJ mm ha - 1 h -1 year -1 was verified.The authors also confirmed an average erosivity of 8,985.46MJ mm ha -1 h -1 year -1 evaluating the regions of Aragarças, Jataí, Rio Verde, Formosa, Pirenópolis, Ipameri, Posse, Goiás, Itumbiara, Goiânia, Catalão, and Distrito Federal.This average value is about 6% greater than the one found for the hydrographic basin of Formoso River (8,483.11MJ mm ha -1 h -1 year -1 ).Carvalho et al. (2017), in the study about susceptibility and current potential to the soil erosion of the soils in the hydrographic basin of Cabaçal River (in the state of Mato Grosso -Brazil), observed that the areas of high potential to erosion of the basin are associated with Neosols, mainly Quartzarenic Neosols.According to the authors, these soils present a very high erodibility degree, and when associated with the incorrect use provide to the area a high potential for soil erosion, as observed in the basin of Formoso River.In another study about eh erosive susceptibility of the soil in the municipality of Paraíso das Águas (in the state of Mato Grosso do Sul -Brazil) through the application of the USLE, greater erodibility results associated with Quartzarenic Neosols were verified, which, according to the authors, for having sandy textures, they are more susceptible to soil loss (Barbosa et al., 2015).Also, Pereira and Cabral (2021), when evaluating soil loss in the upper course of the watersheds of the Taquaruçu Grande and Taquaruçuzinho streams, Palmas (TO), found that the Dystrophic Litholic Neosol was the most susceptible to water erosion with a value of 0.049 t h MJ -1 mm -1 .
The great content of clay and organic matter of dystrophic Red Latosols offer more resistance to the erosion process.Furthermore, these soils have an effective high depth and well developed, which benefits water infiltration and retards the outflow (Lense et al., 2019).In the hydrographic basin of Formoso River predominant soil classified dystrophic Red Latosol of clayey or Very clayey texture, which had the erodibility factor with the smallest potentiality to the erosive processes (0.0131 t h MJ -1 mm -1 ).Consequently, the locations with less presence of this soil have low indexes of soil loss.According to EMBRAPA (2021), the low amount of water available for the plants, and the susceptibility to compaction are some limitations to the use of this kind of soil.
In this sense, Martins et al. (2021) when estimating soil loss in the Córrego Santa Vera watershed on an Oxisol, found that in general, despite the basin being suitable for a tolerant range of soil loss (< 13.3 Mg ha-1 year-1) It is recommended to maintain the riparian forest, in order to avoid silting up of water bodies.
On the other, in locations with dystrophic Red Latosol with medium texture, about 10.59% of the total area presented a very low erodibility class.Moreover presented a greater concentration of real erosion values in the Extremely high class (Figure 9), which can be explained by the compilation of high value of the Factors K and LS in cultivated land areas.According to Sousa and Lobato (2020), the Latosols of medium texture, with high levels of sand, are similar to Quartzite Sands, which are bound to erosion, as presented in Figure 10 (surface erosion), making conservationist practices and cautious management necessaryin clayey Latosols, the attention with erosion is also fundamental.
Figure 10.Degraded areas with the presence of surface erosion in dystrophic Red Latosol with medium texture located in the hydrographic basin of Formoso River, Southwest microregion of Goiás state -Brazil.
Source: Images of 2019 enabled by Google through Google Earth Pro app and organized by the authors (2021).
In the hydrographic basin of Formoso River, we've observed that one of the soil categories with less potentiality to soil loss is Gleisol.In the study developed by Morais and Sales (2017) about the estimative of the natural potential of soil erosion in the hydrographic basin of Alto Gurguéia (in the state of Piauí -Brazil), the authors affirmed that among the main soil classes observed with low and very low susceptibility to erosion, the Gleisols stand out, especially the ones found in riverheads, normally associated with the humid environment of the veredas.
While analyzing the locations with greater results for the LS Factor, it was possible no notice that they are located in areas near the drainage, which coincides with the areas of greater potential and real erosion values.Barbosa et al. (2015) noticed that the erosive dynamics suffer influence mainly from the characteristics of the relief demonstrated through the LS factor, with the predominance of low values (as in the hydrographic basin of Formoso River), consequently, a great part of the municipality didn't present elevated erosion values.The authors emphasize that the areas which present the highest erosion values, classified as strong and very strong, are the same as the greatest values of this factor.In the study accomplished by Alves et al., (2022) in the basin of Verdinho River, also located in the state of Goiás -Brazil, the greatest results for potential soil losses are related, overall, with areas that present greater erodibility of the soils (Neosol texture clayey, and Nitosoil), and greater values for the LS factor.
The results related to the land use and cover indicate the predominance of field formations (62.94%), since the greatest part of the area of the basin is located at PNE. Consideration the part of the area that doesn't belong to the conservation unit, only 24.33% in the basin is composed by natural vegetation, being with 74.12% occupied by cultivated land.The factors which had led to the predominance of cultivated land will be discussed afterward.
The Brazilian grassland formation, in general, is formed by grasses, herbs and small shrubs (Cassino and Ledru, 2021;Silveira et al. 2022) interspersed in the shrub-herbaceous stratum), campo limpid (insignificant evidence of shrubs and subshrubs) and campo rupestre (differentiated from campo-dirty and campolimpio due to the substrate having the composition of rock outcrops and the floristic to include several endemisms) (Overbeck et al 2022;Rossatto and Franco, 2023).
Evaluating the dynamics of the land use in Goiás state among the years of 1985 and 2017, Lopes et al. (2020) verified that, among the years of 1986 and 1996, the suppression of the native vegetation was concomitantly to the expansion of the pasture areas.From 1997, the state has gone through great advancements in cultivated land regarding pasture areas.This information endorses the percentage of the land occupied by cultivated land (74.12%), when considering the area of the preservation unit.
Occupying about 25% of the Brazilian territory, the Cerrado is the second greatest biographic region of South America, is considered the most biodiverse savanna formation of the world (INPE, 2021).Therefore, Parente et al. (2021) reinforce the importance of Brazil in the global environmental agenda, as the necessity of creating legal devices and formulating policies and environmental instruments as a subsidy to an improved and sustainable territory management.For the Cerrado, the Brazilian government created the Project of Monitoring and Deforestation of the Cerrado (PRODES Cerrado), which is the agency responsible for monitoring deforestation by clear cuts and producing annual fees for deforestation in the region (INPE, 2021).PRODES Cerrado carried out the first mapping of the entire anthropic area of the Cerrado biome in 2000 and used that year as a reference, in order to systematically identify all subsequent anthropic increments (new deforestation) from then on, on a biannual basis.(between 2000 and 2012) and annual (after 2013) (Parente et al. 2021).
The CP Factor varies according to the predisposition of the soils to maintain or decrease the potential erosion, with expressive differences between the soil use.In the study, the focus was the cultivated land use, which represents 29.12% of the total area, but it corresponds to the greatest par to the potential erosion Strong and Very strong classes, while all the other uses combined correspond to minor areas for Strong and Very strong classes.Moreover, cultivated land loses 97.78% of all soil annually (real erosion) in the study area.This advancement in the agricultural frontier to inland Brazil occurs since 1970 due to the implementation of policies that intensify land use changes, there's an estimative of a significant parcel, about 45% of the native vegetation of Cerrado, has been removed and converted to other uses, especially cultivated land and pastures (Alencar et al., 2020).
The result causes concern, once the soil erosion decreases the cultivable area, and the productivity of agricultural lands, a fact that must be considered, since the study basin is located in a highlighted region of Brazilian agribusiness, and the poor management of the soils can generate environmental and socioeconomic losses.According to Sartori et al. (2019), in Brazil more than 70% of the arable lands are suffering from severe erosion (> 11 t ha -1 year -1 ) by water, worldwide the losses with the reduction of the estimated global value (annual) reach US$8 billion for the global GDP and the corresponding impact in food safety, with the decrease of agri-food global production in 33.7 million tons.
As in the basin of Formoso River, at which the use and cover had a strong influence on the erosion results, in the study made by Lima et al. (2018), the authors concluded that the alteration of land use and cover characterize an important factor of soil loss, which, different uses present different capacities of intensification of the erosive potential.The authors emphasize that the substitution of the remaining vegetation by pasture or cultivated land represents an addition to the average soil loss of 51% and 110%, respectively.In another study, the authors also found that the most significant soil losses occurred in regions with a predominance of cultivated areas (Cunha et al., 2022).
The destruction of the habitat, the disintegration, and the creation of new barriers from changes in the land use and cover affect the ecological connectivity of different species (Mimet et al., 2016).Therefore, the land use and cover changes must take into consideration the possible negative ecological impacts, and stop them or mitigate them (Almenar et al., 2019).The anthropic actions intensified the erosion in different locations of the study area (Figure 6 and Figure 9), the greatest values were found at agricultural areas near water bodies, which can influence negatively the quality of the water due to the soil particles transport.Barros et al. (2018), in the study about hydrographic basins of Lontra River and Manoel Alves Pequeno River (in the state of Tocantins -Brazil), it was also observed the alteration of the land use and cover in the evaluated period reflected directly on the erosion rates.
In assessing current and future changes in land use and land cover on soil erosion in the Rio de la Plata basin in Brazil, Cunha et al. (2022), found that the absence of soil and water conservation practices caused serious erosion problems in the basin.Still, the riparian forest, although preserved, was inefficient in protecting the watercourse, showing that it is essential to adopt the best management practices in the agricultural production areas of the basin, especially where the ramps are extensive or the slope is greater than 2 %, to reduce the runoff velocity and control the movement of sediments on the surface towards the drainage channels.Kopittke et al. (2019), in the study of the influence of the soil and the intensification of agriculture for global food safety, verified that the soils supply to humans 98.8% of our food.Therefore, the increased growth of the world population has been creating enormous pressure on the soils.To assure global food safety, it's necessary intense agricultural exploitation, which has led to great soil degradation in an unsustainable way.
A study accomplished about the soil loss of the basin of Abóboras Creek, and the basin Laje Creek, also located in the state of Goiás -Brazil, the soil loss varied from 0 t ha -1 year -1 a 732.76 t ha -1 year -1 , and 0 t ha -1 year -1 to 574.37 t ha -1 year -1 , respectively.Among the present use of both basins, the ones which were related to a greater soil loss were exposed soil, cultivated land, eucalyptus, and pasture (Nunes, 2020).A similar result to the ones obtained for the hydrographic basin of Formoso River, at which the greatest values found in cultivated land areas, with expressive results also for pasture, that even representing only 0.11% of the total area contribute with the third greatest value of soil loss in tons.Oliveira et al. (2023) when evaluating the estimate of water erosion in the Jacuba-Goiás river basin (Brazil) through modeling and geospatial intelligence, verified that weak erosion potential with about 97.23% of the area with soil losses ≤ 400 t ha -1 year -1 .The largest area of the hydrographic basin (70.92%) is covered by vegetation of Campo Dirty (44.09%) and Cerrado stricto sensu (23.98%), which indicates that the type of land use directly influences soil loss.
Evaluating the soil erodibility with the objective to contain the erosion of the basin of Wadi Sahouat, Northwest of Argelia, Toubal et al. (2018) verified soil losses estimated between 0 and 255 t ha -1 year -1 .This result is way less than the one found in the present study.This fact may have occurred due to differences in land use between the two basins.The Brazilian basin, when not considered the preservation unit area, is predominantly agricultural.
Moreover, it's worth emphasizing that areas of native vegetation converted for other uses are enough to maintain a high agricultural production in the long run, since proper crops management techniques are applied, with no need for deforestation (Santos et al., 2020).The deforestation threatening the water security and national energy, because the Cerrado is the home of the river sources of the basins of Paraná River, São Francisco and Araguaia-Tocantins, and it also supplies over 50% of hydroelectric power in Brazil (Sano et al., 2019).
The important features concerning the use of Remote Sensing in this study were due to the greater ease in obtaining relief and land cover data, mainly through the use of SRTM and Sentinel 2 images, playing a key role in relief modeling and land use and cover classification.And, about the quality of the image classification, confusions in the confusion matrix were observed (Table 9) among the classes cerradão, stricto sensu Cerrado, field formation, riparian forest, cultivated land.The degree of confusion is associated with the classes spectrally similar that result in the decrease of the Kappa Index, a conflict also identified in the studies by Orozco Filho (2017); Alves et al. (2021); Castro et al. (2022).
Moreover, the difficulty in separating the features may happen due to the barely perceptive transition between the classes (Mercier et al., 2019), the climatic seasonality in the region which influences the behavior of the vegetation, and may also be related to the spatial resolution of the image, justifying the use of images with better spatial resolutions (Nunes and Roig, 2015).
In this sense, the study of soil loss, using the USLE, in hydrographic basins is essential to contribute to the environmental management of hydrographic basins, as evidenced in the study by Weiler et al. (2021), who showed that it is possible to classify areas as "suitable" and "not suitable" for a given use, allowing this organization strategy to identify, quantify and spatialize the areas in accordance with the potential soil loss limit and point out those that do not tolerate tested use, useful information for decision makers when conducting regional planning studies.

Conclusion
The use of geotechnologies along the USLE model enables the obtention of important results over the soil loss from surface erosion in the hydrographic basin of Formoso River.The real annual loss of soil varied from 0 t ha -1 year -1 to 8,315 t ha -1 year -1 , with 88.99% of the total area in the categories Slight to Moderate, and 11.01% in the categories Moderate -High (10 -15 t ha -1 year - 1 ) to Extremely high.
The greatest erosion values are related to areas with the presence of soil classified dystrophic Red Latosol with medium texture, high values of the LS Factor and the use of cultivated land, being areas of higher values close to water bodies.A worrying fact, once, if there's no proper land management, the soil particles may be carried to the rivers, causing problems as eutrophication, turbidity increase, water contamination with pesticides, and, consequently, the unbalance in the water ecosystem.Another fact that can cause a negative impact is cultivated land which surrounds Emas National Park belonging to the study basin, what can affect negatively this Preservation Unit, causing the transport of soils, pesticides, and fertilizers to the areas of the referred park, and, consequently, unbalancing the fauna and flora, besides the contamination of water bodies, what may affect the biodiversity.
The fast conversion of the vegetation cover in different land uses portrays the impasse between the environmental and agricultural policies which determine the fragmentation of the space with the loss of biodiversity or the maintenance of native areas, pondering between the economic development, and environmental preservation.The conflict brings high costs to the ecosystem, as the potential and real soil loss, and any proposal do conciliate the local preservation with the production of food must take into consideration an integrated environment that englobes the geological factors, relief, climate, and soil, but also the human activities, mainly the ones related to the land use and cover, management, and conservationist practices of the soil.
Thus, to avoid greater degradations of the land, this study will be able of being used as an aid to the planning and environmental management of this important hydrographic basin, which is a refuge for the fauna and flora of the Cerrado biome (with worldwide importance), besides other uses.

Figure 1 .
Figure 1.Location map of the hydrographic basin of Formoso River, Southwest microregion of Goiás state -Brazil.Source: Elaborated by the authors (2021) from the data from SIEG's website (2021), organized through the Geographic Coordinates System, Datum SIRGAS 2000 and 22S Zone.

Figure 2 .
Figure 2. Historical average of mininum, maximum, and average montlhy rainfall (mm) for the study área, from 1984 to 2018.Source: Elaborated by the authors (2021) from the data from the rain station no.01852001 Formoso Farm (ANA, 2019).

Figure 3 .
Figure 3.Total and annual average rainfall (mm) for the study area, from 1984 to 2018.Source: Graphic elaborated by the authors (2021) from the data from the rain station no.01852001 Formoso Farm (ANA, 2019).

Figure 4 .
Figure 4. Spatialization of the rainfall stations used in the research.Source: Elaborated by the authors (2021) in UTM projection, Datum SIRGAS 2000 and 22S Zone.

Figure 6 .
Figure 6.Potential erosion in the hydrographic basin of Formoso River, Southwest microregion of the state of Goiás -Brazil.Source: Elaborated by the authors (2021) in the UTM projection, Datum SIRGAS 2000 and 22S Zone.

Figure 8 .
Figure 8. Factor of the land use and cover, management, and conservationist practices of the soil (CP) in the hydrographic basin of Formoso River, Southwest microregion of Goiás state -Brazil Source: Elaborated by the authors (2021) in the UTM projection, Datum SIRGAS 2000 and 22S Zone.

Figure 9 .
Figure 9. Real erosion in the hydrographic basin of Formoso River, Southwest microregion of Goiás state -Brazil.Source: Elaborated by the authors (2021) in the UTM projection, Datum SIRGAS 2000 and 22S Zone.

Table 1 .
Location and period of data collecting in rainfall stations used in the research.

Table 3 .
Soils classification due to the erodibility factor (K).

Table 4 .
Potential erosion classification of soil.

Table 5 .
Real erosion classification of soil.

Table 6 .
Obtained erosivity for each rainfall station.Elaborated by the authors from data enabled by ANA (2019).

Table 7 .
Soil classes and their respective indexes of erodibility of the hydrographic basin of Formoso River, Southwest microregion of the states of Goiás -Brazil.Soils data elaborated from the Soil Map of the Master Plan of the Hydrographi Basin of Paranaíba River (SIEG, 2017), updated according to Santos et al. (2018), ¹World Reference Base for Soil Resources (WRB) (Embrapa, 2021), sistema universal reconhecido pela International Union of Soil Science (IUSS) e FAO.The erodibility values from a Lima et al. (2016); b Demarchi and Zimback (2014); c Corrêa et al. (2015).

Table 8 .
Potential soil erosion in the hydrographic basin of Formoso River, Southwest microregion of the state of Goiás -Brazil.

Table 9 .
Confusion matrix (%) of the image classification of the hydrographic basin of Formoso River, Southwest microregion of the state of Goiás -Brazil.

Table 10 .
Land use and cover, and the CP factor in the hydrographic basin of Formoso River, Southwest microregion of the state of Goiás -Brazil.Elaborated by the authors (2021) based on papers quoted in this table.

Table 11 .
Land use and cover in the categories of potential erosion in the hydrographic basin of Formoso River, Southwest microregion of the state of Goiás -Brazil.

Table 12 .
Real soil erosion in the hydrographic basin of Formoso River, Southwest microregion of Goiás state -Brazil.Soil loss (t ha -1 year -1 )

Table 13 .
Real soil loss by category of land use and cover in the hydrographic basin of Formoso River, Southwest microregion of Goiás state -Brazil.