1 INTRODUCTION
Aiming to regulate the use and conservation of soil, to combat soil erosion, São Paulo State Secretariat for Agriculture and Supply, through the Coordination of Agricultural Defense (CDA), is responsible for applying the Law on Use, Conservation and Preservation of Agricultural Soil - State Law nº. 6.171 / 88 - (São Paulo, State, 1988) to monitor and discipline the soil use and conservation to fight soil erosion and has been carrying out this work for 20 years with very positive results, mainly in inspections carried out at Watersheds.
The area of the state occupied by agriculture has approximately 18 million hectares, with 330 thousand agricultural properties. During that period, 772,000 hectares were already worked at 19,846 agricultural properties that were notified and rehabilitated agroecologically (Vischi Filho et al., 2019).
The CDA Diagnostic methodology uses conservation law as a tool and considers the watershed as the ideal work unit for carrying out this type of activity. The watersheds have an important function of regulating the water balance, as well as housing agricultural production and promoting the storage of rainwater, which seeps into the soil and are available to rivers throughout the year (Lal, 1994; Vischi Filho et al., 2018 and 2019).
Degradation of the watershed is related to the lack or deficiency of vegetal cover of cultures that occupy the soil and has consequently the erosive processes that cause the silting of the watershed water network, interfering in the quantity and quality of the water (Rodrigues et al., 2015). The first principle of soil conservation is soil cover, whether vegetable or mulch. The use of practices, such as the use of varieties that provide bigger vegetal coverage of the soil, reducing the direct impact of raindrops on the soil surface, reduces the soil losses, water, organic matter, and nutrients because of water erosion (Silva et al., 2005; Rodrigues et al., 2015; Merten et al., 2016).
The inappropriate use of agricultural soils causes the gradual loss of its productive capacity and the contamination of water resources by sediments, resulting from the erosion process (Araujo, et al., 2007; Lelis & Calijuri, 2010). The last decades have been characterized by drastic changes in land use and occupation in the region, which for Zalidis et al. (2002) represented one of the main driving forces for environmental degradation, especially on soil and water. This occurs because of the inexistence, or of the erroneous adoption of conservationist practices in the cultivation areas of agricultural properties, a fact commonly verified in Brazil, mainly in pasture areas (Menezes et al., 2009).
Irrational soil management makes production unfeasible and compromises the balance of ecosystems and soil cover (Santos et al., 2007), which is a form of management for a specific crop, in a management system and specific locations, represent a joint effect in reducing water erosion (Silva et al., 2009).
The second principle of soil conservation is to avoid that the surface run-off regime goes from laminar to turbulent and, for that, the construction of an agricultural terracing system is carried out, which has the function of sectioning the length of the ramp and promoting the infiltration of soil water. According to Pruski (2006), the more the soil surface is protected by vegetation cover, against the rain action, lower the propensity for erosion to occur. Studies by Minella et al. (2007) to identify the origin of sediments at watersheds concluded that the areas of crops are the main sources of sediments and suggested that programs for the implementation of conservationist management of soil practices are essential. By adopting proper management and mitigating actions to recover the impacted areas, there will be an improvement in water quality at the watersheds (Araújo et al., 2009). As a final product of soil conservation, its contribution to minimizing floods during periods of heavy rainfall and increased availability of water in the dry period of the year is considered.
In this work on the inspection of Rio do Peixe watershed, the type of intervention aimed at transforming conventional and soil-degrading agriculture into conservationist agriculture, implementing conservationist technical projects that contemplated this novelty. Conservation Agriculture is an agricultural system that promotes the maintenance of permanent soil cover, minimal soil disturbance or no-tillage and the diversification of plant species. It increases biodiversity and natural biological processes below and on the soil surface, which contributes to increasing the efficiency use of water and nutrients, to improve and sustain agricultural production (FAO, 2019).
In 2019, it turned twelve years of activities to inspect the use and conservation of soil at Rio do Peixe watershed, in sections I, II and III, located in Vera Cruz, Ocauçu and Marília cities. This work aimed to test innovations for diagnosis of agricultural properties, to locate erosions and to correct them with changes in the ways of soil management, aiming at transforming the degraded agricultural properties at Rio do Peixe watershed into rehabilitated properties, promoting conservationist agriculture and evaluating the results through remote sensing and water quality indicators.

2 MATERIALS AND METHODS

This work was carried out in the sections of Rio do Peixe Watershed, located in Vera Cruz, Ocauçu and Marília cities, SP, Brazil, at coordinates S22° 14’52.68 ”, W49° 44’59.97”, start and end at coordinates S22° 18’13.28 ”, W50° 2’54.22”, Datum WGS 84 (Figure 1).
The climate of the region is humid subtropical of Cwa type, according to Köppen-Geiger classification, having temperatures in the warmest month above 29.7°C and the coldest month, below 10.6°C, average annual rainfall of 1,193 mm.
Predominant soils are Red-Yellow Ultisol Abrupt, moderate horizon A, sandy/medium texture and Litolic Entisol, eutrophic (Santos et al., 2018). Geological formation consists of rocks from the Bauru Group, covered by neocenozoic sediments (Bezerra et al., 2009). Predominant relief is the smooth undulating, in the western plateau of São Paulo and, in the depression, strongly undulating in the escarpments (Itambé) that separate the plateau from the depression.
CDA inspection/diagnosis methodology (conventional) was developed from 1999 on, by the technical staff of Agronomist Engineers, published in 2003 and it was improved in 2017, receiving the name of Innovative CDA Diagnostic Methodology (Vischi Filho et al., 2019). This happened because action strategies needed to be created to streamline all of this demand. Several options for technological innovations were tested, including the use of model aircraft, helicopter, and drone; however, the results were only favourable when this new method of work was developed. The CDA Innovative pilot project was carried out at Rio do Peixe Watershed, on a 53 thousand hectare stretch, located in Vera Cruz, Ocauçu and Marília cities.
Inspection and agro-environmental rehabilitation work at Rio do Peixe Watershed started on June 15, 2007, and it has been currently happening (from 2007 to 2020). Inspections are carried out at all properties that compose these three sectors of Watershed. CDA Conventional methodology was applied, which consists of preparing an action with the delimitation of Watershed in a topographic map scale 1:50.000, elaborating a mosaic of aerial images, which will be used in the field, “in loco” to help in the localization strategy and, also in the visit to all Watershed properties. Subsequently, the properties are visited using the ”checklist” (official CDA document for Soils), in which erosions and other forms of degradation are included, in compliance with conservation legislation (São Paulo State, 1988).
Erosions are georeferenced using a Global Positioning System (GPS) receiver, classified, photographed (photos that will compose a photographic report) and noted in the ”checklist” leading to the data generated in the diagnosis that will be used to compose the processes relevant to each agricultural property visited and based on the information contained in the diagnosis database. The owner of the area hires an Agronomist Engineer who will prepare a technical conservation project for the recovery of the degraded area, respecting the projects, the class of land use capacity (Lepsch et al., 2015), for this property.
The conservation project is analyzed by the CDA Staff, who approves it or not, sending it for corrections if necessary (physical project, on paper) and if it is approved, it goes to the implantation stage, by the farmer, correcting and remedying soil damage and erosion at that location. The CDA Staff, in this work, is composed of four Agronomist Engineers who work for a week, monthly. They visited each property two to five times during the implementation of the project, the first time during the diagnosis, monitoring the execution of the technical project, and on the property release after the works. After the project was implemented, the soil was preserved and the property became more profitable, increasing productivity. This methodology was applied to two sections of this Watershed, defined as, section I, located in Vera Cruz, SP, Brazil, with an area of ​​14,076 ha (period of work from 2007 to 2011) and section II, located in Ocauçu, SP, Brazil, with an area of ​​9,027 ha (from 2011 to 2015).
In this CDA activity, as well as in other works distributed throughout the territory of the State of São Paulo, having difficulty of traversing the 330 thousand agricultural properties in the state. A strategy needed to be developed and a new methodology for inspection was adopted, it was the Innovated CDA Methodology, which was applied in Watershed section III, in Marília with an area of ​​27,775 ha (from 2017 to 2018).
The methodology consists of making the diagnosis in the office using the databases of Rural Environmental Registry (CAR, 2017), opening and saving the georeferenced “shapefile” with the property limit. To learn about the owner and property information, we used data from the Animal and Plant Defence Management System (GEDAVE, 2017). The CAR and GEDAVE information are specialized in Google Earth® Pro, (current aerial images), promoting an interface of this information with databases, performing diagnosis and inspection by remote sensing. Within the property perimeter, whose shapefile was imported into Google Earth®, the diagnosis begins visualizing the erosions, insertion of a georeferenced GPS point on this erosion, drawing a polygon of this erosion contour, classifying the erosion according to soil conservation legislation (São Paulo State, 1988). Also, elaborating an Excel® spreadsheet containing the following information: the number of georeferenced GPS points, the type of erosion, description of it (Table 2), and erosion area (ha). This work is carried out throughout the property’s perimeter.
After the diagnosis by remote sensing is completed, a colour aerial image is saved in JPEG format and spreadsheet in Excel®, and they are made available to the CDA Staff who will visit the property and go straight to the erosion site, as it is geo-referenced, not needing to be looking for erosion. It is checked whether the erosion of that point exists or not and if it exists, whether it is according to what was described for that point in the Excel® spreadsheet. If the erosions are according to the data in the spreadsheet, the information is maintained and if they are divergent, this information is corrected on the spreadsheet. After visiting all points on the spreadsheet, obtaining photos of the erosions, we have real data on the soil situation of the agricultural property. The data is placed in documents that will be handed in to the notified owner. After this stage, the procedures are identical to the conventional CDA methodology: the owner of the area hires an Agronomist Engineer who will prepare a technical conservation project to recover the degraded area.
To validate the new methodology, some indicators were used to evaluate the results obtained with the implementation of soil conservation. Through the vegetation cover improvement, resulting from changes in soil and water management practices that were proven by comparing the state of the art (before evaluation/work - T1) with the results obtained (after the implementation of technical projects - T2). Using Google Earth® Pro images, through the historic images tool (years: 2002, 2006, 2012, 2013, 2017 and 2018) to evaluate the post-agro-environmental rehabilitation of properties (Figure 2).
The measurement of soil losses and sediment input to the river, due to the action of soil erosion, were evaluated by the water quality indicators evaluated by the Turbidity analysed by the Standard Methods for the Examination of Water and Wastewater - 2130 B method (SMWW a). Suspended Solids were analysed according to the 2540 D method (SMWW b). The Phosphorus were analysed according to the United States Environmental Protection Agency, EPA 6010D method (USEPA). Organic Carbon were analysed according to 5310 C method (SMWW c), which was measured from periodic analyses of Rio do Peixe water.
The samples were obtained from the watercourse, at a station located downstream from the areas where the watershed has the highest concentrations of cultivated areas, and they are collected by São Paulo Environmental Company (CETESB, 2020), once a month, in February, October, and December. According to method 1060 and 9060 of Standard Methods for the Examination of Water and Wastewater (SMWW d), at the collection point code 00SP21438PEIX02100 / UGRHI 21 (coordinates: S22°18’13.62” - W50°2’53.62”).
These indicators were chosen because turbidity shows the sediment input in the water body due to erosion and the transport of these particles to it. The months were defined, as there is a higher probability of erosion, according to rainfall data, considered as the highest rainfall averages, the months of February, October, and December, which were the months that sampled and analysed the water. To subsidize the turbidity and suspended solids assessments in water, the information of Setzer (1985) has used as a comparison.
The data were obtained in two periods, considered as treatments, being: T1 - data referring to the period called BEFORE the working diagnosis and erosion control, considering the information from 2000 to 2007 and T2 - data referring to the period named AFTER the working diagnosis and erosion control, considering the information from 2008 to 2018. The improvement in water quality was assessed by determining the indicators that were tabulated and compared through graphics prepared for each indicator for the T1 and T2 treatments (Figure 3).