Tropical rainforest diversity

Tropical rain forests are the most diverse of terrestrial ecosystems. Many lowland forests contain more than 100 species among the trees of 10 cm dbh or over on 1 ha (Fig. 1.2), and in some more than 200 species may

Figure 1.2 Box-and-whisker plot of species richness for trees greater than 10 cm dbh of rain-forest sites for the major tropical regions (Asia-Pacific, America and Africa). The line inside the box represents the mean value of the average number of species per hectare. The box extends for the range of 50% of the values above and below the mean. The whisker covers the complete range of the recorded values. Data from compilations by Phillips et al. (1994) and Turner (2001).

Figure 1.2 Box-and-whisker plot of species richness for trees greater than 10 cm dbh of rain-forest sites for the major tropical regions (Asia-Pacific, America and Africa). The line inside the box represents the mean value of the average number of species per hectare. The box extends for the range of 50% of the values above and below the mean. The whisker covers the complete range of the recorded values. Data from compilations by Phillips et al. (1994) and Turner (2001).

be found. Species richness rises very rapidly with area or number of individuals sampled in a forest (Fig. 1.3) and plots of 5 ha or more may be required to sample local diversity adequately. Topographic and edaphic variations will often lead to landscapes of patches of different forest communities that further add to the high diversity of a lowland tropical region. The high diversity of species within a particular forest frequently involves the coexistence of species in the same genus. For instance, of the 814 species recorded in 50 ha of forest at Pasoh, Malaysia, by Manokaran et al. (1992), 82% of the species had a congeneric present in the plot, with 70% having a congeneric in the same broad height category.

A commonly asked question, for which we are still seeking the answer, is: why are the tropical rain forests so diverse? Various responses have been put forward. One solution is to turn the question around. Why are extra-tropical regions so poor in species? There is clearly a general inverse relation between environmental harshness and diversity. The tropics have many species because it is easier to survive there than in less favourable environments.

However, this does not explain how all the tree species in the rain forest manage to co-exist. Classic ecological theory states that species can only co-exist if the levels of interspecific competition remain low enough to prevent competitive exclusion of some members of the community. One way in which this can be brought about is for all the species to occupy separate niches. Where do all the niches come from for the co-existence of hundreds of tree species on small areas of lowland tropical rain forest? Various mechanisms have been put forward by which tree species could be partitioning the environment and would thus exist in an equilibrium community. The light gradient from the shaded forest floor to the sunny canopy-top may allow specialisation in different heights at maturity (Kohyama 1993). The horizontal variation in light availability due to irregularity in canopy structure (presence of gaps, etc.) and in soil physical and chemical properties could also result in the subtle environmental variations needed for providing the many niches required.

However, the lowland tropical rain forest generally contains such a large number of species of very similar ecology that it is difficult to believe that strict niche partitioning is occurring. Another possibility is that factors other than niche partitioning may prevent competitive exclusion of species from the community. Compensatory mortality that places a ceiling on the local abundance of a species has been invoked as a means of species co-existence (Janzen 1970; Connell 1971). Species-specific causes of mortality, such as specialist seed and seedling predators and pathogenic diseases, may control the distribution of adult population density within the forest. Recent studies from Barro Colorado Island in Panama showed evidence for significant intraspecific density-dependent effects on recruitment for 67 out of 84 of the commonest tree species in the community (Wills et al. 1997). Givnish (1999) has argued that the fairly strong correlation between tropical rain-forest tree species diversity and total rainfall at a site is an indication of the increasing importance of compensatory mortality as dry spells become more infrequent in the local climate. The increasing rarity of periods of low rainfall allows invertebrates and plant-pathogenic organisms to maintain high population numbers and to provide the mechanism of compensatory mortality continuously.

Alternatively, the species in the forest may not have come to competitive equilibrium. Chance is seen as the major influence on community structure in this non-equilibrium view. There is growing evidence that recruitment limitation is common among species in the rain forest (Hubbell et al. 1999). That is, most species do not establish recruits successfully in all the sites in the forest that they are capable of occupying, either through failure in dispersal or through high mortality of juvenile stages. In a very species-rich community most species are so infrequent that no given pair of species will meet often enough for one to dislodge the other from the community through competi-

Area Tropical Forests
Figure 1.3 Species-area curves for 50 ha plots at Barro Colorado Island, Panama (closed circles), and Pasoh Forest Reserve, Malaysia (open circles). Trees > 1 cmdbh. Data from Condit et al. (1996b).

tion. Recruitment limitation further increases the likelihood of an individual from a competitively inferior species reaching maturity by default.

In reality it seems likely that, to some extent, both equilibrium and non-equilibrium forces operate in the community simultaneously. The bulk of the diversity comes in the form of rare species that are probably not occupying separate niches. However, there seems to be more predictability in forest community structure than might be expected from a strongly stochastically driven system. Long-term studies of primary forest community composition show only minor variations, not random fluctuations, in species make-up and relative abundance over time (see, for example, Manokaran & Swaine 1994). Succession appears to follow relatively predictable trajectories in similar sites (Terborgh et al. 1996). We remain uncertain of what controls or constrains these processes. Hubbell (1997) has proposed that the main influences of chance are on the composition of the regional species pool and the relative abundance of species in the pool. These factors, together with environmental tolerances, are then responsible for community composition at any place in the region. Hubbell has produced some remarkably accurate simulations of community composition (dominance-diversity curves) from his model. However, the mechanics of the interface between regional and local community composition have yet to be explained in detail.

Was this article helpful?

0 0

Post a comment