I examined the relations with branches and foliage of a Hinoki-tree. The diameter of a branch “b” was about 1% of its length “B”. Then I named “Bb” Extent-of-Branch. The weight of foliage of a branch “c” was in proportion to its Bb, the relation was Bb/c=2(cm²/g). And the distance between a branch and the top of the stem “l” was in proportion to the length of the branch of a branch group, the relation was B/l=15-60(cm/m). There was the upper limit of B. In the distance from the limit, the weight of foliage of branches per Δl was in proportion to l. Therefore the crown shape was supposed to be a cone. As a result, the weight of foliage of branches limited branches’ length and diameter. Furthermore the stem shape without its upper part was described with the other parameters. Finally I got a shape model for Hinoki trees.

Goal programming is one of the most effective means for making watershed management plan considering various forest functions. It cannot, however, deal such functions that do not have enough quantitative information because formulation regarding such functions is difficult. We showed how to deal such functions using landscape concept with GIS to make watershed management plan. We clarified the landscape level at which such functions conflict with others and omitted the corresponding area from whole watershed management area, then we used goal programming. When it was difficult to specify the corresponding area, we showed two other ways. One was omitting the area of the upper landscape level that contain the area having probability to conflict and can be specified from whole watershed management area before using goal programming. Another was allocating cutting area using goal programming at first and overlaying the area of the upper landscape level. This makes the area of the landscape elements where conflicts may occur smaller, and we can investigate the whole area to detect where conflicts may occur in it. We applied these procedures to the River Iwaobetu in Shiretoko peninsula and made two watershed management plans.

The relationship between root biomass and stand and site factors were examined using the root biomass data provided by previous reports. It was shown that tree species, stand age, and parent materials strongly correlated with root biomass. On the other hand, root biomass were influenced little by soil type and stand site. Estimation of root biomass showed a significant amount and indicated its considerable role in the functions of carbon storage in forests. Thus, root biomass as well as aboveground biomass, the mass of deposited organic matter, and the amount of soil organic matter are very important factors in the carbon storage of forest resources.

A feasibility study with a prototype model for kiln dried lumber production was performed. The result shows that it is necessary to cut numbers of cost categories for sustainable production of kiln dried lumber. It results as not feasible to invest in wood driers because of low economic efficiency of the Japanese small scale sawmills. It was suggested that diversifying is needed to promote efficiency.

Observing the variation of lumber demand in Japan for 45 years from 1955 to 1999, both the amount of demand for lumber and that for saw log show a maximum value in 1973 followed by a decreasing tendency after that. On the other side, the amount of demand for the pulp and chip material varied differently from the above things to increase until 1995. When the relation with the Japanese economic index is examined, a correlation is seen between the demand for the pulp chip and Japanese economic index, but no correlation is seen between amounts of other demand for lumber and Japanese economic index. To raise the utility of lumber, the concept for the consumption of the lumber must be reconsidered, and then the general consideration of the lumber circulation should be reexamined and expanded. It is suggested that we have to introduce a marketing technique to arouse the consumer’s needs for the lumber.

With an increasing concern about environmental conservation and timber trade issues, results from quantitative analyses have been often provided at such a international meeting as WTO. The objective of this paper is to construct demand and supply functions for imports and exports of industrial logs, lumber, wood panel, chips and particles in Japan, Asian developing nations, North America, Central America, South America, Eastern and Western Europe, Oceania, Russia, and Africa. The derived functions can be used to develop a global timber trade model. The annual data from 1970 to 1999 are used for estimation. We utilize a log-linear model with such an estimation method as the least squares method (OLS), two stage least squares method (2SLS), and three stage least squares method (3SLS) as well as the Almon lag model in order to consider dynamics of economic environments. A non-stationary test is conducted for the data used. Our results show that if parameters estimated by OLS cannot satisfy a sign test, neither can the others. Estimates of supply functions for imports and exports in most cases are not satisfactory in this sense. Only when the amount of imports and exports is increasing or decreasing, the derived parameters show relatively price elastic phenomena.

In this research Timber Supply-Demand Model, which developed for estimation of domestic timber supply and demand under the particular scenarios, is taken up to apply the modified Gentan probability theory to reflect the changes of economic factors such as timber price. As a result, to avoid the complexity of manageable forest problem it is required to adopt the function g(t) satisfying g(0) = 0 and g(t)→∞ when t→∞. The timber price model with monotonously increasing linear trends based on the time series model is proposed as proper g(t) and the parameters of g(t) are estimated by the ordinary least squares method, the maximum likelihood method and so on.

This paper first discusses the censoring and truncation occurred in the data from forest resource table in estimating Gentan Probability. Based on the discussion, a method of maximum likelihood to estimate Gentan Probability is developed, and the properties of the methods introduced by Suzuki and Blandon are reconsidered. Both Blandon’s method and the newly developed method are suitable for data with the censoring and truncation and performed well in a simulation. But Suzuki’s method appeared to have problems and performed worse. Although it has been said that the estimation of Gentan Probability shouldn’t depend on the age structure at the beginning of the observation period, it is concluded that the dependence itself is not a problem if censoring and truncation are well taken into account.