This study examined several aspects of Ester Boserup’s model (1965), which has been highly influential in human ecological discussions on the dynamics of agricultural systems. The starting point of her model is that population pressure is the engine for changes in agricultural systems. For the cultivation stages prior to annual cropping, the result of the mathematical formulation devised in the present study is expressed by a simple equation whose independent variable is the population density in the territory of a community and dependent variable is the duration of fallow period of the fields in the territory. This equation can be graphically depicted as an equilateral hyperbola, although a few other factors may somewhat modify this basic pattern. This suggests that the fallow periods of the agricultural system of a community should be drastically shortened in the early phase of its population growth. At the next step, using a reasonably selected collection of available quantitative data originally obtained from upland rice swiddens in Southeast Asia, the assertion of the model that the labour productivity of a cultivation system with a longer fallow period is superior to that of a cultivation system with a shorter fallow period is tested. On the whole, the results do not confirm this assertion, most likely because such a test is strongly affected by local factors than by the common factor of the fallow period. Still, when the whole data of labour productivity are divided into those of the eight and the nine cases in the equatorial and the monsoonal zones, respectively, the results do roughly show the expected trend that labour productivity decreases with a shortened fallow period. Furthermore, if the two outliers of the nine values of the latter zone are excluded, this trend becomes highly noticeable and statistically significant.

抄録全体を表示

The occurrence and anatomical features of growth rings were investigated in tropical tree species in Peninsular Malaysia. The species studied included five species from two plantations, in a tropical monsoon and tropical rainforest climate, and 26 species from a natural forest in a tropical rainforest climate. Most of the trees were diffuse-porous species, with the exception of Peronema canescens, which was ring-porous. The growth rings were observed macroscopically and microscopically and were classified as having fiber zones, radially flattened fibers, marginal parenchyma, thick-walled fibers, or variations in vessel size and density. Whereas P. canescens from two plantations and Intsia palembanica from the natural forest showed well-defined growth rings, possibly formed annually following environmental stresses, 17 species showed indefinable growth rings and 10 species had no growth rings. Many of the samples presented intermittent growth rings, ambiguous growth rings, discontinuous growth rings, or resin canals and/or traumatic canals that made them difficult to classify. Bands of axial parenchyma were observed macroscopically in Sindora wallichii and Azadirachta excelsa, but microscopic analysis revealed that these bands were formed by concentric traumatic canals. Parenchyma bands with resin ducts or traumatic canals were likely not to be formed periodically. We concluded that some tree species that grow in a tropical rainforest climate do form growth rings and may be used in tree-ring analysis studies.

抄録全体を表示

Mound-building termites function as ecological engineers in tropical savanna landscapes. Large termite mounds support a greater diversity of plants than off-mound areas, but little is known about the process by which diverse mound vegetation develops. This study examined the dispersal mode of woody plants on termite mounds and assessed the process of plant assemblage formation on mounds. This study focused on mound status (termites active vs. inactive) and mound microtopography (mound cones vs. pediments) to develop a chronology of vegetation development. The frequencies of occurrence of woody plants were lower on active mounds (67 %), especially on cones (46 %) compared to inactive mounds (95 %). Species richness and the abundance of woody plants increased in the order of cones, pediments of active mounds and inactive mounds. The proportion of bird-dispersed plant species was much higher on mounds (>40 %) than in off-mound areas (3 %). Salvadora persica was the main bird-dispersed plant and occurred preferentially on active mounds. A schematic plant succession model incorporating the process of endozoochory was developed to explain vegetation development on active mounds.

抄録全体を表示