The influence of species distribution, stochastic events and scalar effects upon species richness was studied within Gambubal State Forest, a subtropical rainforest in southeast Queensland, Australia. The mechanisms restricting species distribution and the occurrence of stochastic events are important in the dynamics of rainforests. On a higher level, the scale of a study influences the way distributions are interpreted and, ultimately, determines the status of each species within the rainforest and hence the species diversity of a community. The work reported is an investigation of some of the limits to distribution and species richness at Gambubal resulting from population level interactions and stochastic impacts.
The distributions of individuals of tree species were examined for four arbitrary size class categories using mapped Cartesian coordinates of each individual. Individuals were categorised as trees, large saplings, small saplings and seedlings. A total of 3004 trees (≥10 cm DBH) were mapped in a rectangular plot of 3.75 ha, 2219 large saplings (1-10 cm DBH) were mapped over 2 ha and 4478 small saplings (50 cm height to 1 cm DBH) over 1 ha.
There was a general tendency for small individuals to exhibit an aggregated distribution that, with increasing size, tended towards randomness. At the small sapling size class, common species did not appear to suffer as a result of high densities. Despite this there was evidence that the small saplings were intolerant of conspecific neighbours as, despite aggregations of conspecific small saplings, the nearest neighbour(s) was likely to be non-conspecific. Amongst the small saplings there were many strong positive interspecific associations found, sometimes multiple positive associations, indicating that there may have been establishment sites that were generally favorable.
Over a two year period, 1082 seedlings were recorded within a 200 m long strip transect. Recruitment was seasonal following flowering and Suiting, with mortality occurring more diffusely over the year although peaking after large recruitment events. Common tree species were also commonly encountered as seedlings. Seedlings clustered near to conspecific trees with low densities away from the immediate proximity of conspecific trees. Higher seedling density near conspecific trees did not appear to exacerbate herbivory or reduced longevity. Herbivory damage was found on 46% of seedlings but such damage did not correlate with seedling longevity over the study period. Litter was the important mortality factor with many seedlings suffering significantly higher litter load preceding mortality.
The distribution of small saplings relative to gaps was also investigated. Many species whose occurrences were only weakly correlated with linear distance to gaps showed a much stronger association with the diffuse light / gap environment. Although these species were positively associated with a gap environment, most species were also encountered in the understorey.
The findings in this thesis suggest show that although the highest rate of change occurs in the smallest size class, population distribution is still changing in the larger sizeclasses. It is proposed that fi-equent small scale stochastic events, such as single leafdrops, have a large impact on newly germinated seedling populations. Because growth is slow in the understorey, individuals tend to remain at sizes at which small-scale stochastic events may result in mortality. Due to the frequency of such small-scale stochastic events, the forest floor under the closed canopy is, in general, best described as sparsely populated by seedlings.
At the scales investigated there did not seem to be a single point or size class critical to the determination of tree distribution, and hence diversity, across the site. Important events were identified (viz. litterfall, development of recruitment space, gap formation) but these were size class specific and were not clearly identified as being pivotal in determining tree community diversity. Rather there was a gradual increase in diversity with increasing size class.