Summary Amazonian forest is seen as the archetype of pristine forests, untouched by humans, but this romantic view is far from reality. In recent years, there is increasing evidence of long and extensive landscape modification by humans. Processes of permanent inhabitation, expansion and retreat of human populations have not always been obvious in those ecosystems, leaving sometimes weak and overlooked imprints in the landscape. An example of one of these inconspicuous alterations are the modifications in the soil known as Amazonian Dark Earths (ADE) or ‘terra preta’ (black earth in Portuguese), which are the product of the accumulation of residuals from permanent or semi-permanent human inhabitation. They are named after the black color of the soils, which is a consequence of the accumulation of charcoal pieces and organic matter in the soil. These soils also contain higher levels of phosphorous, calcium (mainly originated from bone residuals), and nitrogen that increase fertility of the naturally poor soils, thus favouring agricultural practices. Amazonian Dark Earths are distributed in Brazil, Bolivia, Colombia, and Peru, and it is estimated that they could occupy 3% of the area of the Amazon basin. With the decrease in human population in the Americas after the encounter with European colonists, sites where ADE had been formed were abandoned and the vegetation recovered. So far, the effects of ADE on old growth forest had not been widely examined and we are just starting to understand the consequences of past human inhabitation on forest composition and structure. In my thesis, I evaluated the effects of ADE on the forest that has re-grown after abandonment by indigenous people in the La Chonta forest, situated at the southern edge of the Amazon basin, in Bolivia. First, I assessed the magnitude of the changes in the soil as a consequence of human occupation. Then, I studied how soil changes affected plant species composition in the forest understory, forest structure and forest dynamics, and finally I determined how seedlings of tree species respond to anthropogenic changes in soil properties. Detailed information on soil characteristics and its heterogeneity in the landscape is needed to evaluate the effects of soil on the vegetation. Soil heterogeneity in some sites in the Amazon basin can be increased by the presence of ADE. Therefore, I did detailed soil surveys that allowed me to understand the relationship between past human occupation and alteration in the concentration of soil nutrients. I found that natural soils in the southern Amazonian forest are more fertile than their Central and Eastern Amazon counterparts. Past human presence in the area resulted in soil enrichment, due to increases in the concentration of phosphorus, calcium, potassium, and increases in soil pH. Thus, with this information I could test specific hypothesis about the effects of soil fertility on the vegetation that occurs in these sites. In the Amazonian forest in general, soil characteristics influences the composition of understory angiosperm herbs, ferns and palm species. Thus, increases in soil fertility in ADE could affect the distribution of understory angiosperm herbs, ferns and palm species. I evaluated the effect of ADE on composition, richness and abundance of understory species (ferns, angiosperm herbs, and palms). I correlated soil variables associated with ADE, such as Ca, P, and soil pH, with species composition, richness and abundance. I found that the presence of ADE created a gradient in soil nutrients and pH, which changed the composition of understory species, especially of ferns and palms. Additionally, the higher nutrient concentration and the more neutral pH on ADE soils were associated with a decrease in the richness of fern species. I therefore conclude that the current composition of the understory community in La Chonta is a reflection of past human modification of the soil. Soil heterogeneity drives forest structure and forest dynamics across the Amazon region, but at a local scale the role of soils on forest dynamics is not well understood. The study of Amazonian Dark Earths (ADE) opens an opportunity to test how increases in soil fertility could affect forest structure and dynamics at local scales. I evaluated the effect of ADE on forest attributes, such standing basal area, tree liana infestation and successional composition, defined by the relative presence of pioneers, to shade tolerant species in the forest. I also evaluated the effect of ADE on individual components of forest dynamics: basal area growth, recruitment, and mortality. Surprisingly, I found that these fertile ADE affected only few forest attributes and components of forest dynamics. Soil pH was one of the edaphic variables that significantly explained forest structure and dynamics. A higher soil pH increased recruitment of intermediate-sized trees (with stem diameter between 20 and 40 cm) and decreased mortality of large trees (stem diameter > 40 cm). The most important effect of pH, however, was on initial basal area and successional composition, which directly affected growth in basal area of intermediate-sized trees. Increases in soil nutrients can drive plant responses promoting higher growth rates and lower mortality. Plants respond to soil nutrient availability through a suite of traits, by adjusting their biomass allocation patterns, morphology, tissue chemistry and physiology, which allow them successful establishment and regeneration. The higher amount of nutrients found on ADE compared to natural soils could improve the growth of tropical tree species. I studied the effect of ADE on seedling growth, morphology and physiology in a greenhouse experiment with seedlings of 17 tree species from La Chonta. I found that seedlings did not invest more in roots in non-ADE (to take up scarce soil resources) but they invested in leaves and leaf area in ADE (to enhance light capture), although this did not lead to faster growth rate. Tree species responded differently to an increase in soil Ca concentration, which was 2.4 times higher in ADE than in non-ADE soils. Some species seemed to suffer from Ca toxicity as indicated by higher seedling mortality on ADE; others suffered from nutrient imbalance; whereas other species increased their leaf Ca, P and N concentrations in ADE. Only for this latter group of nutrient accumulators, there was a positive relationship between leaf Ca concentration and the growth rates of seedlings. Contrary to expectations, ADE did not lead to increased seedling growth. The ability of plants to colonize patches of ADE might depend on plant responses to increased soil Ca and their capacity to regulate internal tissue calcium to balance nutrition. In summary, in this southern Amazon forest the increased soil nutrient concentrations are a legacy of the humans that inhabited the area. This nutrient addition caused changes in understory species composition and decreased fern species richness and had modest effects on forest structure and dynamics. Increases in nutrients, specifically Ca, can cause positive and negative responses of tree species, resulting in potentially long term effects on the tree species composition of the forest.