Because of the importance of the Amazon to our shared human future and because we need to understand how its forests regenerate, I set out seeds for a week in igapó, palm, terra firme, várzea and white sand forests and then collected them, scoring seed losses to predators, seed losses to pathogens and seeds that germinated. I found (1) terra firme forest, white sand forest, várzea forest and igapó forest under water 1 month every year, were significantly different for seed mechanisms and tolerances, terra firme forest, palm forest, várzea forest and igapó forest under water 1 month per year, were significantly different among species, and the interaction term was significant for all forests except for the two most flooded igapó forests, (2) in terra firme forest seed predators took most seeds regardless of species, (3) in palm forest species were different regardless of seed mechanism and tolerance, (4) in white sand forest seed predators took most seeds regardless of species, (5) in várzea forest seed predators took most seeds but with some species differences and (6) in igapó forest under water 1 month per year, there were differences in predation, pathogens and germination, and in species variation. I conclude that seed predation losses strength as forests become more stressed either by loss of soil fertility or by flooding with nutrient-poor water. Conversely seed pathogens become more important with water-logged soils and with flooding. Seed loss variation among species within forests was always a secondary factor.
Past inability to come to a consensus about the degree of functional redundancy in ecosystems may be due, in part, to different definitions of ecosystem function and different investigative methodologies. Here I define ecosystem function, using the largely plant-based functions of aboveground productivity and decomposition of 10 common early successional trees found in Puerto Rico, and then use two different multivariate techniques to define functional groups. I found that: (1) multivariate statistical methods worked well to sort out the test species on axes defined primarily by productivity, which may have more redundancy than decomposition, and initial leaf nutrient content, (2) there were three plant functional groups defined by species (i) Psychotria berteriana, (ii) Cecropia schreberiana and Inga vera, and (iii) a group containing the other seven species, and (3) the plant traits of nitrogen-fixing capacity and mycorrhizal strategy mapped better onto these groups then those of seed size, wood density, shade tolerance or successional status. Finally, implications for key plant structures and for conservation of Neotropical areas are discussed.
Because decomposition is an important decomposition process and species control of it is an issue in the tropics, we examined how much plant species controlled decomposition compared to environmental conditions on a landslide in Puerto Rico. We chose a landslide because of its extreme special variation in environmental coditions, and found that the landslide center microsite had the greatest variation in temperature and precipitation. Litterbags were filled with single-species leaves of three different plant species and placed in center, border and forest landslide microsites. Bags were collected after 1, 2, 4, 8 and 16 weeks and analysed for organic matter and various other chemicals. All chemicals showed strong differences among plant species and very little significant environmental variation and, with one exception, organic matter loss followed the same exponential decay pattern for all species and microsites. Interestingly, nitrogen (N) and phosphorus (P) immobilization was seen for all three litter types, and calcium immobilization was seen for two of the three. However, immobilization of N and P were most pronounced for the early successional species. Finally, the dominance of species effects over environmental effects suggest that leaf litter chemistry strongly influences decomposition in landslides. While the soil decomposer biota act similarly along landslide environmental ranges, they are more responsive to substrate quality.