Tropical Soil Fertility Changes Under Monocultures and Successional Communities of Different Structure.

For 5 yr we monitored the fertility of a volcanic-ash derived Inceptisol at a site in the humid tropics of Costa Rica. After forest felling and burning, we established four treatments in a randomized block design with six blocks: a sequence of monocultures (two crops of maize [Zea mays] followed by cassava [Manihot esculenta], then the tree species Cordia alliodora), successional vegetation, a mimic of successional vegetation that was physiognomically similar to the model but shared no species with it, and a species-enriched version of successional vegetation. In addition, one plot was maintained free of vegetation. Species-rich successional vegetation was effective at maintaining soil fertility, although we observed general trends of soil-nutrient decline beneath all treatments, presumably because of plant uptake. It proved possible to imitate the fertility-maintaining characteristics of successional vegetation by creating an equally species-rich community of different floristic composition, but the maintenance of fertility was not enhanced by further species enrichment. Successive peaks of nitrate-nitrogen in soil solution, extractable phosphorus, and extractable potassium occurred during the 1st yr, perhaps driven by an early increment of organic matter from postburn debris and roots. Organic matter, total nitrogen, and extractable sulfur were remarkably stable during the 5-yr period. Depletions of cations, decreases in effective cation exchange capacity (CECe ), and increases in acid saturation were related to treatment in the following order: bare soil > monocultures > the three diverse, successional communities. In the bare-soil plot, fertility decreased dramatically: there was a net loss of exchangeable cations and inorganic nitrogen, the phosphorus-fixation capacity increased, and acid saturation reached a potentially toxic 86%. At the start of the study, three of the blocks had soil with lower pH, lower CECe , and higher acid saturation. During the study this less fertile soil lost proportionally more cations and increased more in acid saturation and phosphorus-fixation capacity. The less fertile soil under monocultures proved exceptionally vulnerable to loss of fertility; after 5 yr under monocultures, for example, acid saturation reached 38% in the more fertile soil and 75% in the less fertile soil. In the species-rich communities, however, changes in soil fertility were far less marked.