Seasonally flooded forests represent a transition between terrestrial and aquatic ecosystems. The Mapire river, a tributary of the Orinoco river, floods its surrounding forests during the wet season (May–December). The soils are very acid and the total nitrogen concentration (0.1%) is only half that found in nearby soils flooded by Orinoco waters. Ammonium-nitrogen predominates in the soil during the flooded period while nitrate-nitrogen concentrations are higher in the dry period. Wide fluctuations in the inorganic nitrogen fractions did not considerably affect the annual course of soil nitrogen.

The predominance of mineralization versus nitrification (56 and 5 μg soil month−1 respectively) and possibly the synchronization of nitrogen availability with plant demand could be considered as nitrogen conserving mechanisms.

In synchrony with the hydrologic cycle, the seasonally flooded forest studied shows a nitrogencycle where inputs and accumulation are maximized when the system is under minimum stress (dry season). During flooding, the system enters a period of dormancy making minimal use of nutrient and energy to avoid or tolerate anaerobiosis.

Bioassays of nutrient limitation were carried out for Melastoma malabathricum growing on soil from primary lowland dipterocarp rainforest in Singapore, and for Dillenia suffruticosa on soil from adinandra belukar, a nutrient-poor secondary forest type dominated by Adinandra dumosa. Three questions were addressed. 1. What is the nutrient most limiting to growth in primary forest? 2. What is the nature of nutrient limitation under conditions of adequate P supply? 3. Is there a qualitative difference in the nature of nutrient limitation under primary forest and adinandra belukar? Results showed that there was a strong limitation by P availability in both primary forest and adinandra belukar under the experimental conditions used. Once plants had an adequate P supply, all other nutrients became limiting to growth in primary forest soil. These findings are interpreted as support for the hypothesis that P availability would limit the productivity of moist tropical forests in general in the absence of mycorrhizas; tentative conclusions are drawn on the assumption that most woody tropical plants are mycorrhizal. It is argued that limitation by major cations may be common on old, highly leached tropical rainforest soils.

The population size structure and adult sex ratio were recorded for four indigenous and one introduced species of terrestrial isopod from southern Africa. Interspecific variation was considerable with either discrete or continuous distributions indicative of the production of separate cohorts or continuous recruitment. Intraspecific variation was also considerable particularly in species such as Aphiloscia vilis which can be found in diverse habitats. Sex ratios were consistently female biased, a result consistent with observations made on temperate species. These observations, and a consideration of sexual dimorphism based on body mass, suggest that phenotypic plasticity may be an important tactic in the life histories of tropical woodlice and that in some populations the potential exists for strong sexual competition and complex mating systems.

The populations of two dominant isopod species in a Pinus merkusii forest on the southern slope of mount Ungaran in central Java, Indonesia, were periodically sampled during one year using Tullgren extractions from soil cores. The aim of the research was to collect more information about the life-cycles of tropical isopods by analysing size-class distributions of two sympatric isopod species, belonging to the genus Burmoniscus. Biomass of the populations was estimated from length-weight conversions established for each species in the laboratory. The efficiency of the Tullgren extraction method was estimated by comparison to hand sorting; no difference was found between these methods.

The abundance of the two species differed significantly with time. Burmoniscus Sp. A was more abundant during the wet season, whereas the population fluctuations of Burmoniscus Sp. B were more spread over the year. Reproduction of both species took place in the wet season. Burmoniscus Sp. A had one generation per year, while for Burmoniscus Sp. B the situation was unclear due to extensive overlap of size classes. Comparisons were made with isopod abundance in other tropical and temperate forests. The mean density of the two species found in this study was 258 m-2 for Burmoniscus Sp. A and 272 m-2 for Burmoniscus Sp. B, while biomass was 80.4 mg m-2 and 37.6 mg m-2, respectively. The data suggest that densities of isopods in tropical pine forests may be considerably higher than in comparable temperate ecosystems.

I investigated the extent to which adult butterflies with different wing colours display preferences for flying heights within a lowland tropical forest. Observations on eight colour groups, some of which comprised putative aposematic models and mimics while others contained nonaposematic species, were made from the forest floor and from a scaffold tower that reached above the canopy. Two groups (yellow pierids and papilionids and orange heliconiines) were most frequently observed in or above the canopy, and other groups were most frequently observed in or below the canopy. The evidence suggests that significant differences exist between some groups, but stratification does not occur among all colour groups. In particular, flight distributions of some mimicry rings overlap considerably. The light environments of different forest layers are variable, and the role that butterfly wing colours play in signalling or avoiding prcdation is little understood. However, the general lack of fine stratification of colour groups suggests that flight preferences do not occur because wing colours and patterns are especially cryptic against the background of particular forest layers.

A field experiment was conducted to test the effects of simulated folivory on leaf production, plant growth rate and mortality of small individuals of Metrodorea pubescens in an evergreen gallery forest in central Brazil. Although complete defoliation decreased annual growth rate, both in absolute and relative values, it also stimulated leaf production. Mortality was observed only in individuals in the smallest size class, but was not statistically related to percentage defoliation. Recovery of completely defoliated individuals was rapid and after one year over 70% of their initial number of leaves was present. It is probable that insect folivory is not a direct cause of mortality of juveniles of this species.