COS 17-4
Tropical forest structure and diversity in response to varying disturbance regimes
The effect of disturbance on biodiversity is the subject of on-going debate, with both complimentary and competing hypotheses presented to describe the response of diversity to various disturbance regimes. Tropical forests, with their high biodiversity and rapid species turnover, present ideal conditions for testing hypotheses that relate species richness to disturbance. The Lake Hargy Caldera of West New Britain, Papua New Guinea contains a lowland tropical forest that has been subject to multiple large and small scale disturbances ranging from natural gap fall to catastrophic flooding and volcanic disturbance. This study aims to quantify the response of biodiversity to disturbance by examining: 1) primary production in the various disturbance-prone regions of the caldera, as represented by basal area and stem density of each constituent species, 2) relative α and β-diversity within- and between-areas, and 3) species turnover within areas. We recorded species composition and structural metrics of seven plots within four distinct disturbance-prone areas of the caldera. NMS and MRPP were used to describe differences among all sample communities, and structural indices were applied to explain these differences. Correlations between diversity metrics and structural indices by plot and by species were used to describe differences due to inferred disturbance history.
Results/Conclusions
Between-area plot differences were significant (p=0.011, T=-2.81), while plots within areas were more similar to each other than different (A=0.18). The NMS plot demonstrates significant differences between samples from opposite sides of the caldera, with samples in the center of the caldera clustering in the center of the NMS plot. The subplot with the highest α-diversity also had the highest total basal area, but the highest cumulative basal area by plot pair was correlated with the lowest overall diversity. Results demonstrate local-scale effects of disturbance on productivity within a background forest matrix, and provide support for both dominant-species-mediated diversity (i.e. Mass Ratio Hypothesis) and intermediate disturbance hypotheses, depending on the nature of the disturbance.