Gradients in moisture, soil fertility and elevation constitute the most important variables in tropical forest environments. Changes in any of these environmental attributes are likely to affect the performance of functional groups or to cause shifts in the relative abundance of functional groups within tropical ecosystems.
In tropical regions, annual rainfall varies from near zero in the Atacama Desert in northern Chile, and a few centimeters in the Guajira Peninsula of northern Colombia, to more than 10 m in the upper San Juan Valley of western Colombia and in northeast India. Over much of the Neotropics there is a good correlation between total annual rainfall and the length of the wet season because the amount of rainfall during wet months is relatively constant over broad areas. However, many regions in southeast Asia have extremely high rainfall during the summer monsoon, combined with a long dry season. Other areas have low rainfall fairly evenly distributed over the year (Walter 1973).
A decreasing gradient in moisture is likely to be associated with a higher frequency of fires. Because fires affect functional groups differently, ecosystem processes may be altered accordingly. For example, if understory shrubs are more affected than trees, the resources upon which a large number of pollinators and fruit eaters depend in tropical dry forests may be extensively diminished by fire, altering the energy flow interface at the level of primary consumers. Fires also favor animals able to escape by moving out of the area or by burrowing (Braithwaite 1987).
Production of flowers, fruit and litter is more clumped temporally in regions with long dry seasons than in regions of relatively constant rainfall (Opler et at. 1976; Foster 1982a,b; Lieberman 1982; Leighton and Leighton 1983). Therefore, the How of energy occurs in marked pulses in tropical dry forests, and the temporal concentration of litterfall concentrates patterns of nutrient retention and transfer (Silver et al. 1996). The relative importance of plants that flower and fruit during the driest season is probably also positively correlated with increasing dryness.
When tropical soils become saturated with the first heavy rains following a lengthy dry period, there is often an increased incidence of treefalls (Brokaw 1982; Brandani et al. 1988). Thus, as dry seasons increase in duration, the temporal clumping of treefalls becomes more pronounced. Drying/wetting cycles also accelerate the replenishment of the available soil nitrogen pool from microbial, recalcitrant or physically protected nitrogen pools. Fluctuations in soil moisture cause crashes in populations of soil microbes that induce pulses of nutrient release. These cycles in soil nutrient availability and moisture may increase the uptake of limiting nutrients by plants (Lodge et al. 1994). With increasing length of dry season, the drying of the soil becomes more extreme and the pulsing of microbial populations and nutrient release probably become especially marked, but the influence of this pattern on ecosystcm productivity and the efficiency of natural nutrient cycling is yet to be determined. Although extreme drying of the soil is associated with long dry seasons, even relatively brief rainless intervals can cause large reductions in soil moisture. At La Selva, Costa Rica, a site without a well-marked dry season, there was a 40% reduction in total soil moisture content in the upper 70 cm of soil following a 1-month period without significant rainfall. Such reductions are sufficient to cause water stress in the forest vegetation (Sanford et al. 1994).
Although belief that all tropical soils are red, infertile and harden irreversibly when they arc cleared is widespread, soils of the lowland tropics are as diverse as those of any other region. Infertile red and yellow oxisols and ultisols are common throughout the tropics, but red, infertile soils are found on only about 7% of the tropical landmass (Sanchez 1976), and there are significant areas of highly fertile soils along rivers and in volcanically active areas. A reduction in species richness, vegetation layers, canopy height and mean leaf sizes is associated with decreases in soil fertility in tropical forests (Brunig 1983). Data from the few tropical forests that have been studied intensively enough to provide good comparative information have been summarized by Jordan (1985). The sites range from a forest on a rich dolomite soil high in magnesium- and calcium-carbonate in Darien,
Panama, to forests on infertile spodosols and oxisols at San Carlos, Venezuela. Changes in ecosystem functioning associated with the gradient from high to low soil fertility are listed below.
1. Decreasing production. Productivity of leaf litter ranged from 11.3 Mg ha1 year 1 at Darién to 4.95 Mg ha 1 year"' at San Carlos. Wood production data are not available from the most fertile sites, but the medium-fertility sites had nearly double the production of the least fertile site, suggesting that the range in wood production is comparable to that for leaves.
2. Reduced above-ground biomass but increased below-ground biomass. Standing above-ground biomass varied by a factor of two, whereas below-ground biomass varied by a factor of 10. As a consequence, root-shoot ratios ranged from 0.03 at Darién to 0.49 at San Carlos.
3. Reduced decomposition rates. Trees on fertile soils produce large quantities of nutrient-rich non-scleromorphic foliage that decomposes rapidly. In contrast, trees on infertile soils produce smaller quantities of nutrient-poor, scleromorphic leaves that decompose more slowly. The rate of leaf decomposition varied by nearly a factor of five along the soil fertility gradient.
4. Increased percentage of roots in a superficial mat. 20 25% of the roots in the forests at San Carlos are in a superficial mat; such mats do not exist in forests on fertile soils.
5. An increase in the relative allocation of plant resources to defense (Coley et al. 1985) and a reduced allocation to reproduction (Gentry and Emmons 1987). These differences in resource allocation contribute to the slower decomposition rates of litter in forests on infertile soils, and result in lower populations and richncss of species of animals that depend on pollen, nectar and fruits.
6 A reduction in species-richness, vegetation layers, canopy height and leaf sizes (Brunig 1983).
Associated with increasing elevation on the slopes of tropical mountains are increasing wind, rainfall and water logging of soils, and greater incidences of landslides (Leigh 1975; Lawton and Putz 1988). The following biological changes are correlated with these physical changes.
1. Decreasing productivity. Few data are available for tropical premontane forests, but litter production at montane sites in the Luquillo Experimental Forest in Puerto Rico is only half as great as in productive lowland forests (Odum 1970), and wood production is similar to that in lowland forests on highly infertile soils.
2. Decreased above-ground biomass but increased below-ground biomass. The root mass at EI Verde, Puerto Rico, is greater than that of lowland forests on rich soils by about a factor of six, but above-ground biomass is lower by only a factor of two, in part because the leaf biomass of the premontane forests is similar to that of lowland forests.
3. Reduction in decomposition rates and an increase in litter accumulation. These changes increase nutrient retention in decomposing litter and decrease rates of nutrient transfer. Litter accumulation also dramatically alters the soil surface and the composition of the litter fauna.
4. Lower species richness in most taxa (Terborgh 1977; Janzen 1987).
5. A reduction in the diversity of plant life forms. Canopy height decreases, palms drop out with increasing elevation, and there is an overall reduction in leaf sizes (Leigh 1975; Tanner and Kapos 1982; Brown et al. 1983). in addition, there are shifts in the relative representation of life forms. Vine and lianas become less common and epiphytes increase in abundance and structural diversity (Brown et al. 1983), which increases nutrient capture, nutrient retention and nutrient transfers at the atmosphere-terrestrial interface (Silver et al. 1996).
Was this article helpful?