This study examines the ecological characteristics and processes that lead to the formation of stands dominated by A. thurifera (Dipterocarpaceae) in Papua New Guinea. The studies examined seedling regeneration through to stand development and explored how this ecological information might be used to develop silvicultural methods for managing these forests.
The distribution of A. thurifera across PNG was assessed and the species was found to occur in two situations. In some sites it occupied more than 40 percent of the basal area or 50 percent canopy cover and formed what shall be referred to in this thesis as "monodominant" stands. These stands were common in forests along the Morobe Coast. The second situation when there were found where A. thurifera was less dominant in the stand. The forests at Kuriva and Sagarai typified this.
The impact of logging on forest containing A. thurifera was assessed at five paired logged and unlogged sites. Logging commonly removes 25 - 40 percent of the stand basal area, depending on the original species composition, size class distribution and logging intensity. Logging therefore affects the recruitment and growth of A. thurifera. In certain instances, where more than 70 percent of the basal area is removed, few residuals are retained and the capacity to replenish large disturbed areas is minimized. However, with low intensity logging, species composition of the original forest is retained. In some sites, the effect of logging was found to increase A. thurifera populations over time as well as in increased growth of the residual trees.
The flowering and fruiting patterns of A. thurifera also affect the way regeneration occurs at disturbed sites. A total of 158 trees of varying size classes were monitored over four years at Oomsis forest. Data analysis showed that the timing of flowering varies but most tend to flower at the late dry period and the early wet season, which is between August and November. It also showed that A. thurifera has a more regular flowering and fruiting pattern than many other dipterocarps. Mast fruiting occurs every three to five years but otherwise there is sporadic annual flowering and fruiting. The community flowering and subsequent fruiting lasts a maximum of nine months from the first appearance of flowering to the final seed rain.
Seed rain was assessed using 217 traps of 1.8 m x 2 m laid at the base of parent trees and in a grid pattern to capture the seeds. The data showed that the quantity and quality of seed production depended on parent tree size and the flowering intensity. The most productive size class was stems larger than 50 but < 90 cm dbh. It was further found that majority of the seeds were dispersed within ten metres radius of the parent tree crown.
The viability of A. thurifera seeds was further assessed by examining the size class of parent trees and at different stages of the seed rain period. The results showed that viability increased according to seed maturity and was highest immediately upon seed fall. Further tests to determine seed viability upon storage at room temperature revealed that those stored for 21 days all lost viability but those stored for 14 days had 68 percent viability.
Predation was assessed at different stages of the seed rain. The results showed that predation commences very early during the development phase of the seeds and continued until the end of the fruiting season. It was found that pre-dispersal predation was high affecting 50-65 percent of the total seeds produced in a season. Major predators were insects, birds and pigs. It was also found that the population of the predators increased during the fruiting period, especially of birds (parrots) and pigs. The overall predation was found to be 85 percent of the seeds produced.
The dynamics of A. thurifera seedlings on the forest floor under closed canopies was examined by establishing four 2 m x 2 m plots under each of 23 parent trees over three years. This showed that the number of seedlings regenerating depended on the intensity of fruiting and stand density, averaging at 0.42 seedlings m-2 during the study period. The continued presence of established seedlings on the forest floor depended on factors such as the ability of the species to persist and the forest type in which it occurs. At Oomsis, being a secondary forest, 64 percent of the original seedlings persisted over three years, and 79 percent of those recruited in December 2001 survived after seven months. While mortality was found to be high immediately after germination this stabilised over time. Mortality was also high in the lower height classes and decreased with increasing height classes. Seedling height growth averaged at 7 cm year-1 and the highest growth rates were recorded in younger seedlings and among those growing under higher light levels.
The optimum requirements of sunlight for the growth of A. thurifera seedlings were further examined in a study using five levels of light treatment over 56 weeks. The treatments were, 10, 30, 50, 75 and 100 percent Relative light Intensity (RLI). Results showed that the highest growth rates were attained under the 30 percent RLI. While survival was 100 percent in all treatments, the control had 93 percent survival over the same period. Growth in the first 16-20 weeks were unaffected by the quantity of light, suggesting that this species can germinate and get established under poor light conditions. However, light becomes essential for the advancement in growth. It also showed that open light conditions are unsuitable for the growth of this species.
The effects of silvicultural treatment on the growth, recruitment and mortality were examined at Oomsis in a regrowth forest stand that had developed 20 years after logging. Four treatments were used; high thinning, medium thinning, low thinning and an unthinned control. Treatment prescriptions were BA limits. The results of this study showed that the highest growth rates for A. thurifera were attained under the high thinning treatments. Similar responses were found for other species present in these stands. The abnormally dry weather conditions extending over nine months during the study period and an anomalous experimental design may have affected some of the results. There was no significant effect of the treatment on recruitment or mortality over the previous five-year study period. The results also showed that the trees in the 20 - 40 cm size class responded greatly to the treatments. This are the potential crop trees for the next harvest.
This information was used to explore the silvicultural prescriptions that might be used to manage A. thurifera forests in PNG.