RESUMO
Amazonian Dark Earths (ADEs) are unusually fertile soils characterised by elevated concentrations of microscopic charcoal particles, which confer their distinctive colouration. Frequent occurrences of pre-Columbian artefacts at ADE sites led to their ubiquitous classification as Anthrosols (soils of anthropic origin). However, it remains unclear how indigenous peoples created areas of high fertility in one of the most nutrient-impoverished environments on Earth. Here, we report new data from a well-studied ADE site in the Brazilian Amazon, which compel us to reconsider its anthropic origin. The amounts of phosphorus and calcium-two of the least abundant macronutrients in the region-are orders of magnitude higher in ADE profiles than in the surrounding soil. The elevated levels of phosphorus and calcium, which are often interpreted as evidence of human activity at other sites, correlate spatially with trace elements that indicate exogenous mineral sources rather than in situ deposition. Stable isotope ratios of neodymium, strontium, and radiocarbon activity of microcharcoal particles also indicate exogenous inputs from alluvial deposition of carbon and mineral elements to ADE profiles, beginning several thousands of years before the earliest evidence of soil management for plant cultivation in the region. Our data suggest that indigenous peoples harnessed natural processes of landscape formation, which led to the unique properties of ADEs, but were not responsible for their genesis. If corroborated elsewhere, this hypothesis would transform our understanding of human influence in Amazonia, opening new frontiers for the sustainable use of tropical landscapes going forward.
RESUMO
O objetivo desse trabalho foi avaliar dois métodos de laboratório para a determinação da condutividade hidráulica do solo saturada (Ko) conhecidos como Permeâmetro de carga constante (PCC) e Permeâmetro de carga decrescente (PCD), com o intuito de verificar sua aplicabilidade e variabilidade em solos amazônicos. Coletaram-se 125 amostras de solo com estrutura indeformada, através de amostrador tipo Uhland, com anéis volumétricos, de 0,072 m de altura e 0,069 m de diâmetro, devido à variabilidade apresentada pelas determinações de tal parâmetro. Nos mesmos pontos de amostragens da Ko, procedeu-se coleta de anéis volumétricos para a determinação da porosidade do solo. Ainda nesses pontos foram coletadas amostras com estrutura deformada para análises físicas e químicas. Os resultados obtidos demonstram que o método do PCC foi o mais apropriado para a classe dos Latossolos estudados, apresentando os menores coeficientes de variação e desvio padrão ao longo da topossequência. Os valores de Ko estiveram distribuídos entre P1(2,65 à 3,34 cm dia-1), P2(2,85 à 3,38 cm dia-1), P3(2,86 à 3,63 cm dia-1), P4(2,75 à 3,49 cm dia-1), P5(2,38 à 3,83 cm dia-1) e P6 (2,47 à 3,52 cm dia-1); havendo uma tendência para maiores valores de Ko na superficie. A utilização de Ko como parâmetro de análise hídrica em solos porosos na superfície e muito argilosos em profundidade, como os amazônicos, necessita ser realizada com precaução, evitando a interrupção da continuidade dos poros e compactação da amostra. Mudanças na condutividade hidráulica saturada estiveram mais relacionadas a alterações nas propriedades físicas do solo e posição no relevo do que nas alterações das coberturas vegetais ao longo da topossequência.
The objective of this work was to evaluate two different laboratory methods for determining the saturated hydraulic conductivity (Ko), namely, the constant head permeameter method (PCC) and the falling decreasing head permeameter method (PCD) and their applicability and variability to Amazon soils. 125 undisturbed soil samples were collected with an Uhland soil sampler using volumetric rings of 0,072 m height and 0,069 m in diameter. Soil porosity was also estimated by volumetric ring samples collected at the same spots where Ko was evaluated. Disturbed soil samples were also collected for chemical and particle size analysis at the same spots. The results showed that the PCC method was more appropriate for the studied soils Oxisols, leading to the lowest coefficient of variation and standard deviation throughout the topographic sequence. The Ko values were distributed among P1(2,65 to 3,34 cm day-1), P2(2,85 to 3,38 cm day-1), P3(2,86 to 3,63 cm day-1), P4(2,75 to 3,49 cm day-1), P5(2,38 to 3,83 cm day-1) and P6 (2,47 to 3,52 cm day-1); having a tendency to show higher Ko values at soil surface. The use of Ko as a parameter for hydraulic analysis in soils with high porosity in the surface layer and high clay content in the subsoil, as is the case in the Amazon, must be undertaken with caution to avoid compacting the sample and porous discontinuities. Throughout the studied topographic sequence, changes of saturated hydraulic conductivity were more related to changes in soil physical properties than to changes in vegetation cover throughout the studied topographic sequence.