Conversion du cerrado en pâturages cultivés et fonctionnement des Ferralsols

The latest agricultural frontier of the savannah to be exploited for intensive agricultural use on the south American continent, the Brazilian cerrado and the Llanos of Colombia and Venezuela, make demands on research as more of this land is brought into production. The introduction of exotic Poacea during the last three decades has enabled the livestock carrying capacity to be increased 5-10-fold. However, these systems progressively lose their productive capacity five years after being introduced. During an IRD-EMBRAPA project the functioning of ferralsols has been studied with regard to its physics, hydrodynamics, biology and the geochemistry of phosphorus against a background of the transformation of the natural savannah vegetation into cultivated Bracharia spp pastures. This paper describes the main balances and attempts to clarify the features of the soils under pasture which are affected, and in particular ferralsols, which represent 46% of the area of the biome. Most of the pastures brought into cultivation are on these soils. The work has shown that the state of the soil under an old, unproductive pasture is characterised by medium-low organic matter content and very low concentrations of plant-available phosphate ions. The surface horizon usually exhibits a loss of porosity and the soil structure is altered, being more compacted. The hydraulic properties do not seem to be greatly affected, and runoff and erosion are only localised problems resulting from incorrect soil management practices used on certain watershed catchment basin. The diversity and biomass of the macrofauna are reduced. However the physical and biogeochemical behaviour recorded indicate that properties favourable to plant growth can be restored by changing the management of the system. With the development of agriculture and the introduction of pastures, the cerrado is affected by the big reduction in plant biomass. In the pastures the functional area is not the field, but the plant unit associated with the exactly underlying soil volume. The scale is metric. The biological mechanisms have been modified on two scales: on the macroscopic scale, with the alteration of the invertebrate fauna (e.g. of the parameters of the distribution of communities and population characteristics, i.e. their abundance and biomass) following changes in the amounts and nature of the organic matter entering the system; and on the colloidal scale, since it is observed that the characteristics of enzymatic activity are sensitive to soil transformations. The changes in porosity and structure which have been identified are the result of mechanical effects on the soil and biological changes. Conversely, the microporosity of the soils is conserved, whereas this is not the case with the deforestation of the Amazonian soils. The variations in porosity observed also concern the possible fluctuations in organic matter content. The regulatory mechanisms of the structure and the macroporosity by the soil invertebrate fauna are modified. The hydrodynamic processes at the surface are in fact regulated by the plant cover; we confirm that runoff and particle size detachment may be regarded as relatively unimportant. The hydraulic conductivity varied little, but this needs to be confirmed. The direct effect of physical and soil fauna modifications on the pastures productivity has not been tested. The soil fauna is not responsible for the falling yields: at the very most, there is a synchrony between the ageing of the pasture and the adaptation of the fauna to the food supply in the environment. The porosity, which falls at the surface, is somewhat restricting to the root development in the whole soil volume, but this factor must be added to the herd management factors. On the other hand, we have shown that these soils remain very susceptible to fluctuating rainfall, the available water reserve declining rapidly, particularly when the fine porosity falls. But the cause of the decline in this case is the occurrence of short dry seasons.