Search

We made litter fall observations in a West Sumatran tropical rain forest every 2 weeks through 3 years (including the drought year, 1997). Seasonal changes in long-term average rainfall were bimodal, but even in the two dry seasons average monthly precipitation was > 200 mm. Leaf-fall rates in a year with normal rainfall peaked in the first dry season; leaves made up one-third of mean annual litter fall. Larger trees had highly synchronized leaf fall during the peak shedding period. Smaller trees shed simultaneously during severe drought. Twigs

The guild composition of a hill dipterocarp forest was examined using growth traits of 1566 trees comprising 422 species through field observation at a 6-ha plot for 6 years in Sumatra. Three parameters, intrinsic growth rate, attainable maximum stem diameter, and stem hardness, were used for guild determination. The intrinsic growth rate showed a significant correlation with stem hardness, which suggested a trade-off between volume growth and architectural strength under the restraint of photosynthetic products. Four major guilds were determined, characterized as Soft wood-Small size-Fast growth, Soft-Big-Fast, Hard-Big-Slow, and Hard-Small-Slow traits. Ecological niches of these guilds were related to the large variance of resources in time and space of a tropical rain forest. A secondary forest part in the 6-ha plot retained high species diversity, but species composition was largely different from that of the mature stand. Human activities biased guild composition to reinforce the Soft-Small-Fast guild, with a reduction in the Hard-Big-Slow guild that included many marketable timber trees. Logging impacts on dipterocarp and fagaceous trees were assessed with reference to their spatial distribution and attributes of guilds.

The aggressive reproduction strategy of the naturalized shrub Piper aduncum L., introduced from humid Neotropical regions, was studied in Sumatra paying special attention to fruit production and leaf longevity. A single piper shrub produced inflorescences throughout the year and the annual production of a mature inflorescence was estimated at 7.8 ton ha-1 yr-1. It was also shown to consume 100 kg of nitrogen per hectare for annual fruit production on the basis of its crown projection area, while 60 % of nitrogen in leaves and branches was reabsorbed with senescence and death. Average leaf longevity was 158 ± 48 days with an average turnover rate of 2.5 ± 0.51 yr-1. This high turnover rate is considered advantageous for continuous fruit production from early growth stages through effective allocation and reabsorption of energy and nutrients. Next, a mathematical model for optimum leaf longevity was proposed taking into account the maintenance costs of a single branch. The model showed that the observed leaf longevity approximately satisfied the expected minimum maintenance cost of a mature branch. A short leaf life span could save stem and branch construction costs through effective leaf production with low biomass. Economical branch structures in terms of energy and nutrient use could allow allocation of more resources to fruit production. This prioritization of fruit production over construction of a durable stem and branches could be evaluated as an adaptive living strategy of this shrubby pioneer species. Based on the proposed model, the adaptive survival strategies of other species were discussed with special reference to the relationship between leaf longevity and branch structure.

The altitudinal distribution of Fagaceae in West Sumatra was studied on the basis of herbarium specimens and field tree censuses conducted at Gunung Gadut and Sipisang. Forty-two species were examined (Castanopsis: 8 spp., Quercus: 8 spp., Lithocarpus: 25 spp., and Trigonobalanus verticillata), and it was revealed that altitudinal distribution patterns vary among species and genera. The species richness of Fagaceae in West Sumatra was highest at 400-700 m, decreasing with increasing altitude above 700 m, while about half the total number of species occurred below 400 m. Though the altitudinal distribution patterns of some species were compressed toward a lower altitude in Gunung Gadut, species richness in Gunung Gadut and Sipisang regions were also highest at 400-700 m above sea level (76 % of 33 species were found growing in these regions), and more than half of the species were found at 150-400 m. Quercus showed gradual replacement/overlapping of higher/lower altitudinal species at a range of 700-1100 m, and Lithocarpus showed a remarkably high species richness at 400-700 m in both regions. The altitudinal diversity patterns of Fagaceae appeared to involve both species coexistence/concentration at particular altitudes and species replacement with changing altitudes. The high species richness of Fagaceae at 400-700 m was especially due to the coexistence and concentration of Lithocarpus species.

The altitudinal gradient of vegetation of the Barisan range in a coastal region near Padang, West Sumatra, was studied in 10 plots (0.1-6.55 ha) distributed from 600 to 1780 m. Trees with a diameter at breast height of more than 10 cm were employed for classification of the vegetation. Four vegetation zones, two types of montane and two types of lowland vegetation, were recognized based on cluster analysis, floristic information, and the physiognomy of these forest stands. An upper zone of montane vegetation (Lower montane Fago-Myrtaceous forest) was distinguishable from a lower zone (Sub-montane Fago-Myrtaceous forest) due to the remarkable dominance of several species of Fagaceae and Myrtaceae, and its lower canopy height. The two zones of lowland vegetation (Upper hill dipterocarp forest and Mid-hill dipterocarp forest) were distinguishable by their species composition of Dipterocarpaceae and stratification of the stand. The altitudinal zonation corresponded well with classic zonation patterns previously reported for West Malesia, but the critical threshold altitudes were lower in comparison with those in other regions. The borderline between the lowland and montane vegetation, corresponding to a Kira’s warmth index of WI = 180 °C · month, was around 200 m lower in Padang than in other regions of West Malesia, and the critical altitude between the two zones of lowland vegetation was also around 300 m lower in Padang than on Mt. Kinabalu. The characteristics of altitudinal zonation in Padang were discussed with reference to the climatic features of this region, namely, the low temperature and hyperhumidity.

Leaf longevity of tropical shrub species was examined with reference to the intensity of disturbance. Species with lower stem hardness likely prioritize a higher volume of growth at early growth stages over the physical strength of the stem. This trait is thought to be advantageous for species suffering frequent disturbance, thus it is believed that stem hardness could be used as a measure of adaptability to disturbance. The weight ratio of leaves to stalks on a twig is an important parameter to assess the leaf longevity satisfying the minimum maintenance costs of a twig. In this study, the leaf longevity of shrubs was shown to be reciprocally related to the weight ratio. Based on this relationship, it was revealed that these species gain suitable leaf longevity by regulating this weight ratio under conditions of the minimum maintenance. A positive correlation between leaf longevity and stem hardness was also observed among shrub species. Assessment of the niche of these species in a secondary sere according to stem hardness verified that leaf longevity increases with progressing succession as a result of regulation of the weight ratio.

The floristic composition of a tropical hill forest in Ulu Gadut, West Sumatra, Indonesia, was studied. High species diversity and differences in the floristic composition compared to other forests in West Malesia were observed. Based on the tree numbers of common species in 20 × 20m quadrats, habitat axes representing floristic composition were determined by detrended correspondence analysis. Axis 1 was significantly related to the slope convexity and stand height of a quadrat, indicating that the floristic composition is influenced not only by topography but also by growth stage and disturbance. The major disturbance in this forest prior to the study was local logging, the influence of which was found not only in the variation in floristic composition but also in vegetation development after disturbance. We concluded that the harvest of upper layer trees by local loggers enhanced the growth of lower layer tree species in this forest. The estimated original vegetation without logging differed from that of other forest stands in West Malesia because of its relatively low dominance of Dipterocarpaceae and high dominance of Fagaceae. This might have been caused by the lower temperature of the plot compared with other forests at similar altitudes in West Malesia regions. This climatic feature is also thought to contribute to the high diversity of the plot.

Decomposition processes of leaf litter were observed during one year with same kinds of substrates at a warm-temperate evergreen oak forest in Japan and a tropical rain forest in Sumatra simultaneously. A warm-temperate forest showed high rates of carbon-weight loss in summer. A tropical rain forest had high rates at the initial stages in particular and apparently low rates during the severe dry weather in early 1997. When time trends of carbon-weight loss could be approximated by a single exponential curve for each leaf litter, the average values of decomposition factor of five Japanese fagaceous leaf litter were 0.70±0.13 y-1 at a warm-temperate forest and 1.62±0.24 y-1 at a tropical rain forest, respectively. The C/N ratios of litter gradually decreased with the progress of decomposition at the initial stages, then tended to be constant at about C/N =20 at the last stages irrespective of kinds of leaf litter and study sites. Basing on these results we discussed characteristics of decomposition at these two forest ecosystems.

Successive changes of stand structure and primary productivity of a ropical secondary forest from clear cutting were studied through long-term observation during 12 years from 4 years later after the event in a foothill rain forest area of West SumaEa, Indonesia. Species number of trees over 5 cm dbh at the initial observation ranged from 30 spp. to 40 spp. per a stand being under 1000 m2 in area, and this value was less than half of one in a matuled forest stands with same area. The secondary forest had some prominently dominant species derived from germination after the last logging. Growth rate of biomass was largely different among sites, and a sund on fertile soils conditions grew up to 300 metric ton ha-1 in aboveground biomass within 9 years. Many canopy trees in the forest, however, rapidly died at the latter phase of this observation, and Eees sprouted from logged stumps and species in a matured forest both increased their dominance relatively. Frequent logging would be to the advantage of species having coppice generation because ofpositive correlation between a stump size and growth rate of shoots. Hardness of stem wood of secondary species tended to be lower than one of matured forest species, and the property had a positive conelalion with relative growth rate of dbh (RGRD) for secondary species. A mathematical model based on this relation suggested that amount of photosynthetic product of a tree would not be largely different irrespective of stem hardness under similar external and internal conditions, and a tree would allocate the product to suitable positions for the growth ofvolume or bulk density based on is life srategy. A time rend of a primary net productivity had the maximum value within a few years after logging, and tended to be constant. The value at the stable stages was not sigrificantly different from one of a neighboring matured forest, and was estimated at26 - 30 metric ton ha-1 y-1 as total primary productivity including roots. The RGRD - dbh of a stand largely changed with time on regeneration process, and it produced the maximum productivity at the initial stages. Basing on this time rend of this relation, we analyzed successive changes of stand structure and primary productivity under consideration with effecs of stem hardness of a Eee as physical property.

An equatorial rain forest in West Sumatra suffered defoliation and severe stem damages from two storms in early 199S, Intensity of litter production at a storm generated gap site was 225 ton ha-1, and big woody debris over 10 cm in diameter accounted for 78%. A mass of fine litter promptly lost its weight of dry matter in Ao layer directly after the event, and the Tate tended to have a positive correlation to the initial amount of accumulation. An impact of these mass litter production on a material cycling of an equatorial rain forest ecosystem was simulated basing on these field data and monitoring data of a forest since 1981. The simulation showed that fine litter would supply nutrients enough for revegetation at an initial stage in particular, and a great amount of woody debris would contribute to the process at later stages because of their low decomposition rate. The time needed for 95% loss of initial litter amount by decomposition at a gap site was estimated at 11 years in this forest i[respective of amount, when the composition of litter is same. This average decay rate overcomes a net primary productivity. From a carbon balance between both potential rates of decay and growth for initial 11 years after a damage, it shows that about 100(organic matter) ton ha-1 of litter production is a boundary value between its sink and source for this stand.

Horizontal variance of a stand structure and productivity in a tropical foothill rain forest was analysed with the dbh census data obtained from the successive observation for 8 years at a 1-ha forest stand in West Sumatra, Indonesia. Targets of the census were trees over 8 cm in dbh. Individuals at each dbh class distributed randomly in horizontal pattern. The quadrat method based on the dimensions per some unit area showed that the aboveground biomass was higher aggregated in spatial distribution because of the strong influence of the scattered emergent trees. Main canopy trees being about 35 cm in dbh largely affected the horizontal distribution of leaf biomass and productivity, and the degree of contagion of these values was lower than that of aboveground biomass. We estimated the stand structure and productivity around 2336 points in a plot, arranging systematically at intervals of 2 m, from the values per projection area of a tree covering each point. Bird’s-eye-view diagrams were presented to visualize the horizontal distribution patterns. The three-dimensional distribution of estimated relative illuminance showed the horizontal and vertical stratification in a 1-ha forest stand. The frequency distribution of biomass, productivity and relative illuminance at lower levels under 20 m in height could be approximated by the log-normal distribution. The coefficient of variance of productivity was higher than those of aboveground and leaf biomass. A simulation of the regeneration process at a gap site in a mature forest reproduced the high variance of productivity.