Portal de Periódicos da CAPES

Artigo Acesso aberto Produção Nacional

BIBTEX RIS

Status of Soil Surveys and Demand for Soil Series Descriptions in Brazil

2013; Wiley; Volume: 54; Issue: 2 Linguagem: Inglês

10.2136/sh2013-54-2-gc

2163-2812

I. F. Lepsch,

Agricultural Science and Fertilization

Crop forecasts for this year show that Brazil will produce 180 million metric tons of grain (including 80 million tons of soybeans), which should place the country as the first tropical “agricultural giant” and the first to challenge the supremacy of the “big five” food exporters in the world (United States, Canada, Australia, Argentina, and the European Union). “Even without a good infrastructure of roads, government subsidies, and detailed soil surveys, Brazil is now the third largest global producer of soybeans. When they have all the facilities, as we do on our premises, we will go bankrupt.” “Modern soil science, spearheaded by research in Brazil has facilitated the utilization of vast areas of previously uncultivated soil long considered unsuitable for human food production into highly productive agricultural land. Naturally acid soils with high contents of aluminum and iron oxides, low CEC values, and organic matter contents long considered insurmountable obstacles to crop production in tropical latitudes could be extremely productive. With continued development of the infrastructure needed by commercial agriculture, Brazil has the potential to lead the world in its quest to provide food for growing human populations”(Buol, 3). The remarkable expansion of Brazilian grain production in the last four decades took place mainly in the Central Brazilian tablelands, the so called Cerrado region. The Cerrado biome is, for the most part, geographically located south of the Amazon forest region and accounts for about 25% of the country's land mass. The original vegetation varies from open grasslands to areas with grass and short trees and shrubs, up to more densely placed trees, generally adapted to natural fires. Mean annual temperatures average between 22 and 27°C. Rainfall ranges from 800 to 2,000 mm, concentrated from September to March and followed by about five months of nearly no rainfall. Before 1960, no farming or intensive grazing took place in the Cerrado, and with a very low population density, the land cost was cheap and the main agricultural usage was as extensive pasture for bovines. Progressively, more experienced farmers from south of Brazil moved into the area, bringing commercial farming practices and grain production. The uniform temperatures and the dry season allowed for ideal conditions for sowing and harvesting the grain over extended periods of time, resulting in maximum efficiency of machinery operations and reducing time and cost to dry the harvested grain. In these lands, two cycles of crops per year are possible, even without irrigation if no-till systems—now a common practice—are used. Field experiments testing recommendations for liming and fertilizers were established shortly after the new capital, Brasilia (D.F.), was built in the center of the Cerrado region, in the 1960s. Since the beginning of modern agriculture in this frontier region, farmers always had assistance for soil chemical testing, nitrogen biological fixation techniques, application of agrochemicals, and selection of plant varieties adapted to the local climate conditions. This assistance has been provided by both private and governmental institutions. Igo Fernando Lepsch is a retired soil scientist from Instituto Agronômico de Campinas, SP, Brazil, and a visiting professor at the Universidade Federal Rural do Rio de Janeiro, RJ, Brazil. In the Cerrado, also called “edaphic savannas,” the dominant soils are the deep, low-base, highly weathered Latossolos (mostly classified as Ustox in the U.S. Soil Taxonomy). However, soils were never mapped beyond the exploratory and reconnaissance level (1:1,000,000 and 1:500,000 scale, respectively). Plans for construction of new railways and waterways continue to this date, increasing the infrastructure for agricultural production. However, I feel this is not enough. In order to make the best practical use of the soil information, land managers must know not only the “what” and “why” about the different soils, but also the “where” concerning the location of more economically manageable areas of different soils. Many agronomists contend that higher crop yields could be obtained if the land originally under Cerrado vegetation could be managed better by choosing species and crop varieties adapted to the various soil types. Environmentalists also argue that this increase in productivity per acreage would be the best way to prevent further clearing of the areas that remain under the coverage of the Amazon forest. Unfortunately, with the exception of highly organized sugarcane commercial plantations, little has been done in Brazil to make possible soil mapping on an appropriately larger and detailed scale for agricultural business and small-farm applications. The lack of large-scale soil maps is my biggest concern about the efficiency of soil surveys in Brazil, and the following relates my viewpoint for this status of knowledge about our soils. Following graduation in 1961 from the Federal Rural University of Rio de Janeiro, I went to work for the Soil Conservation Service of the São Paulo State Government. After a training course, modeled from a previous one offered by U.S. soil conservationists, I started to map soils in farms for the planning purposes of introducing soil conservation practices. I mapped soils using coded legends similar to the ones recommended by the first edition of the Soil Survey Manual (Kellogg, 8) that a few years earlier were used to map soils on U.S. farms, before the official establishment of the USDA Soil Conservation Service. Looking back, I now believe that I was mapping the farms at a “soil series level” (1:5,000 scale) even without knowing how to classify the soil pedons in a taxonomic system. This was done by distinguishing in the field the soil bodies that differed in slope shape, slope degree, soil texture, color, depth, drainage, etc. In 1966, the State of São Paulo government closed the Soil Conservation Service section, based on the assumption that the mapping of the farms, using coded legends, for soil conservation was no longer a priority. I transferred to a research institute, the Agronomic Institute of Campinas (IAC), and continued to work in soil science. As part of the IAC staff, I was approved for graduate studies at NCSU, working under the guidance of S.W. Buol and R.B. Daniels in the subject of soil landscape relationships (Lepsch et al., 11). Back in Brazil, after achieving a Ph.D., I realized how important studying the stratigraphy, geomorphology, and hydrology of the landscape was to better understand the genesis of Brazilian soils as well as the usefulness of these relationships to establish criteria for detailed soil mapping procedures. Most of the Brazilian territory is located in very stable crystalline shields and in humid tropical environments, resulting in a landscape without recent influence of volcanoes, earthquakes, or deserts. During the Pleistocene, a series of stable geomorphic surfaces and their correlative superficial deposits were formed. According to geomorphologists, these Brazilian landscapes were sculptured by different processes and under alternate humid and dry morphoclimatic conditions related to the glacial and interglacial periods occurring in the Northern Hemisphere, where the glaciers covered much of the region and many soils had been formed in sedimentary materials left by the massive continental glaciers of relatively young geological age (Holocene). This is a situation very different from that found in Brazil, where, at the beginning of the semiarid phases, the active erosion forces over deeply weathered regolith mantles widened the valleys forming several minor planation surfaces that today are incised by younger Holocene valleys. Within the Guiana and Brazilian shields, the mineral material, from which a great deal of the present soils formed, was pre-weathered and redistributed during many episodes of erosion, especially during the beginning of the semiarid climates coincident with the glacial stages and deposition later on. As a result, the sediments of the surficial deposits contain very few weatherable minerals having potassium, calcium, magnesium, or phosphorus, since they have been dissolved while in previous regoliths and/or during transport (Buol, 3). Most present day soils in those surfaces are Latossolos (well-drained Oxisols) formed from deep sedimentary layers that had been exposed to many cycles of weathering, erosion, and deposition (Lepsch and Buol, 10). Most Latossolos are considered to be polygenetic soils, and the more reworked their parent material was, the more weatherable and devoid of bases the current soils are. As silicon and bases were removed from the regoliths during the cycles of erosion and deposition, kaolinite-type clays and iron and aluminum oxides were formed. The association of clay and iron oxides forms a strong fine and very-fine granular structure, often called “pseudo-sand” (Fig. 1). The result is that very small voids within the granules retain water at high tension, and the large voids between the granules allow water to rapidly percolate under the force of gravity. This accounts for the characteristic of these Latossolos: even with high amounts of clay, they are very permeable and have good physical properties such as good tilth capability. These soils are found adjacent to rejuvenated young slopes formed during the Holocene with less acid and weathered soils (Fig. 2 and 3). All of this points to the importance of the stratigraphic and geomorphic approach to delineate soil mapping units that could be used to “baptize” the Brazilian series and help standardize first- and second-order soil surveys. To achieve this, the new field soil scientists have to be trained to recognize soil profile characteristics as well as landscape, stratigraphic, and hydrologic features, thus mapping the land based on soil landscape conceptual models. The soil bodies from the Cerrado tablelands of Central Brazil have yet to be delineated on first- and second-order soil maps in order to be “baptized” with the series names. On top of the tablelands (called ʺchapadasʺ), the naturally acid Latossolos (well-drained Oxisols) were originally under “cerrado” vegetation and presumed insurmountable obstacles to crop production in tropical latitudes. Nowadays they are highly productive and intensively cropped. The entrenched valleys around the tablelands have less acidic and less weathered younger soils. Photo by Rodrigo E. M. de Almeida. The Brazilian Federal Government program for soil survey implementation began in 1950. At that time, it was imperative to inventory the soils of the country (8.5 million km2), most in the Amazon and Cerrado regions, with very little human occupation. The first scientists to work with Brazilian pedologists were from European institutions, working as FAO/UNESCO advisers. Since the FAO/UNESCO had initiated the world soil map project, there was a great call for soil data covering the large Brazilian territory to compose the South America Soil Map (Beek and Bramão, 2), which comprised the first part of the FAO World Soil Maps. Thus, the priority was for exploratory and reconnaissance soil surveys. These small-scale mapping projects (1:500,000 and 1:1,000,000) were coordinated by the federal agency that now is known as EMBRAPA-SOLOS. In these original soil maps, taxa names were derived primarily from the 1938 United States system (Baldwin et al., 1), plus Kellogg's (9) proposals for tropical soils and the FAO/UNESCO soil map legend (FAO, 5). Names such as Podzólico Vermelho-Amarelo (Red Yellow Podzolic Soils), Latossolo, Litossolo, Nitossolo, Podzol etc. were used in Brazil for many years. In 1980, after much testing of the U.S soil classification system (Soil Survey Staff, 12), Brazilian soil classification began adopting principles and concepts that parallel Soil Taxonomy, but with more emphasis on tropical weathered soils. Another landmark was the Radam Brasil Project that went from 1970 to 1985. This project was dedicated to covering various regions of Brazil (especially in the Amazon Basin) using side-looking radar aerial images captured by airplanes. The use of radar images allowed the soil surface to be reached even under dense cloud cover and forested areas. Interpretation of these images and fieldwork produced a comprehensive integrated study of the physiographical regions and biomes, producing analytical texts and thematic maps on geology, geomorphology, soils, vegetation, land use potential, and renewable natural resource potential. Upon completion of the Radam Brazil Project (which generated soil maps at 1:1,000,000 scale and nationwide), the number of field soil scientists in governmental agencies was reduced. Since then, a comprehensive federal program for systematic mapping, municipality by municipality and in scales compatible with the land-use planning needs of farmers, has yet to be established. Under these circumstances, by 1980, the soil scientists of the EMBRAPA-SOLOS decided to focus on preparing a Brazilian System of Soil Classification (SiBCS) with the help of various Brazilian universities and other research institutions. The SiBCS was first released in 1999, its second edition came out in 2006 (EMBRAPA, 4), and a new edition is planned for this year. A generalized 1:5,000,000 scale soil map of the whole country based on this system was published (IBGE/EMBRAPA, 7). This system recognizes 13 soil classes at the highest categorical level of order. Four categories were already outlined: order, suborder, great group, and subgroup; definitions of families are currently being studied and tested. However, the soil series were not mapped and established to work as a base for the system validation. Many aspects of U.S. Soil Taxonomy were incorporated, plus some classes from the FAO legend and from the World Reference Base of Soil Resources (FAO, 6). The orders are differentiated mainly by the presence or absence of well-defined diagnostic horizons, such as the “horizonte B latossólico” (some similarity to the oxic horizon of Soil Taxonomy). Several other quantitative diagnostic characteristics are considered for the suborders and great groups such as base saturation, clay activity, cation exchange capacity, effective cation exchange capacity, sodium saturation, and some distinctive morphological characteristics. Most subgroups are differentiated according to the taxa central concept for great groups and properties indicating intergradations to other categories or soils with extraordinary characteristics. Different from Soil Taxonomy, neither the soil moisture or temperature regimes are considered in the SiBCS. Unlike U.S. Soil Taxonomy (Soil Survey Staff, 12), which was based on thousands of mapped and registered soil series, the SiBCS developed in a descending order, i.e., considering the population of the first few soil classes—uppermost categories (orders and suborders), which are gradually subdivided and classified into the lower categories (soil great groups and subgroups). The system is being constructed mainly by analyzing data from pedons sampled for the many reconnaissance and exploratory-level Brazilian soil surveys. For the detail desired in the sixth level, the soil series, several arguments have been raised about the precariousness of conditions for their establishment. Despite the problems to manage in setting the soil series in Brazil, the current high demand for soil mapping is in the detailed surveys (first- and second-order surveys) defined by “the series level.” Thus, taxa at the series level must be established. Examples of this level of detailed soil survey are found on lands where farmers grow sugarcane or eucalyptus to provide their harvest to pulp and paper companies and sugarcane and alcohol industries. Since this knowledge is not openly available, there is urgency to conceptualize and define the soil series mapping units. Among the detailed soil surveys undertaken in Brazil (most by “freelance” soil scientists), patterns for mapping and presentation of maps are highly variable. Similar symbols may be used for different soils and similar soils may get different symbols, which makes communication and interpretation of soil data difficult. Discussions about the concepts of “series as taxa” have started among Brazilian pedologists; however, there are no “series mapping units.” Moreover, assuming, as in other countries, that a soil series can only be officially recognized after a minimum area is mapped, in addition to taxonomic standards, it is necessary to discuss and define criteria that must be used for recognition of the mapping units that would be equivalent to the future “official Brazilian soil series.” The SiBCS (EMBRAPA, 4) is a remarkable achievement and has advanced our knowledge about Brazilian soils. This system provided a categorization of Brazilian soils based on well-defined soil profile distinguishing characteristics as well as criteria that dictate choices in use. However, the definitions of the yet undeveloped lowest categorical level that has been done in the SiBCS—order, suborder, great group and suborders—consider soils as single points (soil profiles) in the continuum; only very few statements about landscape features are considered. This soil taxonomic system, like many other national ones, helps to organize our knowledge about our soils. In this way, the properties of classified objects may be remembered and their relationships may be understood most easily. However, since we do not yet have defined categories that most closely correspond to soil bodies—the soil series and their phases that should be grouped into families—it does not help detailed soil surveys much. Also, with an emphasis on profile characteristics, many soil scientists tend to have a tunnel vision, looking only inside the soil pits without integrating them in the landscape. So, there is an urgent need for a national program to train field soil scientists and to provide sufficient resources to produce detailed soil maps and soil survey reports with interpretations of the mapping units for practical purposes. In the USA, initial funding to accelerate soil survey came when Dr. Hugh H. Bennett (during the Depression and Dust Bowl years of the 1930s) convinced Congress to release funds for the Soil Erosion Service that later became part of the USDA-NRCS Soil Survey Division. In the Brazilian scenario, EMBRAPA-SOLOS has only a few dozen field soil scientists, and in order to map about 8.5 million km2 of the yet unknown Brazilian soil series, we require a much larger specialized personnel. Hopefully a catastrophic event like the one mentioned above will not be needed to convince our government to release sufficient funds to initiate a detailed soil survey program in Brazil. The author acknowledges Professor S. W. Buol (North Carolina State University), Mr. John Kelley (USDA-NRCS, retired), Professor Thomas Jot Smyth (Tropical Soils Program Leader at North Carolina State University), and Professor Lúcia Helena C. dos Anjos (Universidade Federal Rural do Rio de Janeiro) for the helpful review and suggestions.