Forest stand variables provide important information for sound forest management. We developed estimation models for forest stand variables such as number of trees per ha (N), stand basal area (BA) and stand volume (V) of undamaged trees in a tropical forest after a fire. We used the spectral bands of SPOT XS and the transformation bands including the normalized difference vegetation index (NDVI) and inverse vegetation index (IVI) as independent variables. The regression technique was used for the model selection. Stand volume and basal area of undamaged trees were found to be significantly correlated with NDVI and IVI. For stand basal area, the best model was BA=22.838-28.982 IVI with R^2=0.729; and for stand volume V=316.707-405.661 IVI with R^2=0.791. Such quantitative forest information may be useful for monitoring the forest.

View full abstract

Approximately 90% of industrial wood worldwide is harvested under concession agreements. Failure to resolve the problems with forest concession operations has encouraged unclear or wrong management decision making, which has resulted in changes to the forest cover. Cambodia was chosen as a case study. The aim of this report was to examine the current problems with forest concession management, which is very important for present and future policy-decision making. Cambodia currently has 10.6 million ha of forest cover; 4.7 million ha of which are managed by forest concessions. Owned by multi-national corporations, forest concessions play a vital role in economic development in Cambodia, whilst at the same time they protect the forests from illegal encroachment. However, illegal logging and over-exploitation of forest resources were regularly reported to have taken place both inside and outside the concession areas. Wood production in Cambodia in 1997 was 3.4 million m^3 (5 times higher than the sustained yield rate), 68% of which came from illegal logging. Various unofficial payments imposed on forest concessionaires and high logging production costs have put logging operations in Cambodia in a position of losing $ 33.43-76.45/m^3 of veneer at the current veneer market price. Additionally, due to the intensive illegal logging, approximately 10%, 50% and 40% of operable forests will be logged out in 10-15 years, 5-10 years, and less than 5 years, respectively. It clearly showed that a forest industry crisis will be likely to occur in the next 15 years from now. The lack of clear land use policy, mapping, boundary demarcation and law enforcement encourage illegal logging and corruption in the Cambodian forestry sector. Therefore, all these aspects must be addressed to bring the remaining forests under sustainable use and management. Since illegal logging activities involve various parties, full cooperation from individual, local, national and international institutions is required. Additionally, a pilot project on forest concession management should be initiated to provide the right direction for future management decision making.

View full abstract

We compared the estimates of forest areas in China from the forest census database (1989-1993) and the land cover database based on remote sensing data of the NOAA AVHRR-derived monthly NDVI composite data from April 1992 to March 1993 at 1-km resolution, and analyzed the spatial distribution of forest resources. According to the land cover database, China has a total forest area of 1.27 million km^2, about 3% lower than the estimate from the forest census database. Our county-level quantitative analysis of forest area and population in China characterized the scarcity and uneven spatial distributions of forest resources per capita in China. On the average, forest area per capita in China is less than one-fifth of the global average of 0.64ha. Forest area per capita in North, Central, and South China is substantially lower than the national average of 0.11ha, and the spatial distribution in Northeast, Central, and Southwest China is relatively uneven. This study also highlights the challenges for using forest resources to improve the ecological environment, to raise people's living standards, and to spur economic growth in China. Based on analytical results, we proposed three perspectives concerning China's forest resource management in the 21st century.

View full abstract

Two fractal dimensions of tree crown images, Ds for silhouettes and Do for outlines, were examined in relation to crown transparency (CT) using 137 standard photographs for the visual assessments of CT in the European forest health monitoring program. The automatic thresholding and the dilation method were applied using NIH Image to estimate Ds and Do. It was found that Ds and Do can not necessarily specify a level of CT, and rather, they were associated with the two components of CT, the foliated area and foliage transparency. A new measure DSO (=Ds-Do) is proposed and tends to decrease curvilinearly with increasing CT, with DSO being close to 0 at CT=100%.

View full abstract

I tested whether or not BEHRE'S and KUNZE'S relative stem taper curves in coniferous species can be estimated theoretically only from tree height and diameter at breast height through a theoretical two-way volume equation, and compared the accuracy and precision of the estimated taper curves. The estimation procedure was fitted to 50 Cryptomeria japonica D. DON and 50 Chamaecyparis obtusa ENDL. trees. The estimated BEHRE'S relative stem taper curve showed higher accuracy and precision than KUNZE'S. In BEHRE'S relative stem taper curve, the mean relative errors of estimated diameter were less than ±3.2% for C. japonica and ±11.8% for C. obtusa, except on the upper stem near the tip. The standard errors of estimated diameter ranged from 0.7cm to 1.2cm for C. japonica and from 0.5cm to 1.9cm for C. obtusa. In conclusion, BEHRE'S relative stem taper curve can be well estimated theoretically only from tree height and diameter at breast height, except on the upper stem near the tip.

View full abstract

The scope of forest planning is rapidly extending beyond traditional timber production to sustainable management of biodiversity. Biodiversity in forestry usually refers to variety of species but also includes genetic variations and spatial variations of landscape. The usual approaches in forest biodiversity quantification, however, face fundamental problems in measurement and analysis, because diversity is a particularly complex and difficult subject at the landscape-level. In this study, we evaluate forest compartments of the Miya River Watershed, Mie Prefecture in Japan with a newly proposed index of landscape biodiversity called the Land Use Diversity Index (LUDI). The results of this study indicated that the LUDI was an effective method both for calculating forest landscape biodiversity and for planning tactical forest management. In particular, by using empirical information on forest vegetation, we were able to detect areas of forest that were most and least important for maintenance of biodiversity in forest planning.

View full abstract