One reason why trees grow differently over the year may be due to their relative spatial condition. In this paper, we introduce a statistical model called the varying coefficient model to estimate effects on the tree growth from the age and relative space perspectives. This model is a superior way to understand the tree growth, which is because 1) the space effects can be added in the growth curve model and visually checked by drawing contours and cubic diagrams, and 2) both the space and the age effects are estimated with their confidence interval values, and also the hypothesis test can be performed, which improve the confidence of the estimation result. The model is applied to the data of Cryptomeria japonica, the longitude and the latitude are used as the space information, and the significant spatial effect of the DBH on the stem volume was evident from the estimation result.
This study aimed to examine the relationship between topographic factors and the photosynthetic rate parameter of a process-based growth model for sugi (Cryptomeria japonica) planted forests. The site-specific photosynthetic parameters were estimated using stand growth data derived from 30 permanent plots established in the Tano Forest Science Station (University of Miyazaki) by Bayesian calibration. The estimated plot-specific photosynthetic rates were related to the following topographic factors: solar radiation index, hydrological upslope contributing area index, vertical and horizontal exposure index, and average slope gradient. Significant but weak correlations were found between the plot-specific photosynthetic rate and all topographic factors except solar radiation index.
The PATH algorithm (Paredes and Brodie,1987) is interpreted by means of the calculus of variations. Using the PATH algorithm, a new dynamic programming model called Stand Optimization System (SOS) is developed. The system is incorporated into a growth simulator constructed by Arney (1985). Further limitation of optimality on the PATH algorithm and the relationship between the Lagrange multiplier and the decision variable are discussed.
This paper examines the importance of income generated from community forest to the rural poor in Kaski District, Nepal. The results of the study show that on average, households earn 7.4% of their cash income from community forests. Poor households are more reliant on forest activities compared with the better off. They earn 13.6% of their total household income from community forest compared to the rich households who earn only 2.1%. The results of the study also reveal that income from community forest have a stronger equalizing effect on local income distribution. The Gini coefficient was computed as 0.37 when income from community forest was considered and 0.53 when it was ignored. These findings show the importance of community forests to the rural poor and underprivileged households. The findings also suggest that in designing community forestry programs, policy makers should not ignore socio-economic disparity among the forest user households.
Forest stands and individual trees are often devastated by natural disasters such as typhoons and heavy snowfall in Japan, resulting in significant economic losses to the forestry sector. Our objective is to identify key risk factors that affect the degree of damage. We apply two types of statistical approach: one is, a logistic regression model to snow damage data to investigate if there is any geographical element affecting the degree of damage at the stand level, and the other is, a survival analysis on tree failure data for factors affecting the degree of damage at the individual tree level. A logistic regression analysis revealed that the risk probability of snow damage is higher on older and thin stands. The analysis also indicates taking advantage of certain geographic conditions to reduce wind burden could decrease the degree of damage. A Cox regression analysis showed that tree age, diameter at breast height, and species were key factors that influenced the degree of tree failure. Specifying risk factors throughout statistical modeling helps to provide a comprehensive, systematic, and objective method to assess risk in forest management.