We examined the errors of slope and aspect calculated from digital elevation models (DEMs) derived from １:25,000-scale topographic maps published by the Geospatial Information Authority of Japan (GSI). First, we generated DEMs with spatial resolutions of１０mfromcontoursat 10-m intervals and spot elevations in “Fundamental geospatial data (level 25000)”generated by GSI and a DEM with a spatial resolution of 10m from 10-m mesh data, the “Fundamental geospatial data (grid altitude information)”generated by GSI. We calculated the slope and aspect angles from the DEMs. Then we established 10×10-m experimental plots corresponding to DEM cells in forests using the global positioning system and a laser range finder, and measured slope and aspect angles with a clinometer at the centres of the plots. Finally, we determined errors in the estimates from the DEMs using the plot measurements. The estimates of slope were not biased, and their root mean square errors (RMSEs) were less than 10 degrees. Excluding sites with gentle slopes, the estimates of aspect were not biased, and their RMSEs were less than40 degrees. However, we note that the estimates of aspect could have large errors at sites with marked changes in slope or aspect such as valley lines and ridge lines, despite wider horizontal contour distance.

Changes in stand structure and growth of a mixed forest managed by selection cutting were analyzed based on long-term monitoring data (57-years) measured at three experimental plots (dominated respectively by broadleaved species, coniferous and broad-leaved species, and coniferous species) in Yubari city, central Hokkaido,northern Japan. As a result of selection cutting management, the tree mortality has decreased, the commercial timber value of the remaining trees has increased, and the growth of the stand volume has increased. However,it was found that in the plots dominated by broad-leaved species and by coniferous and broad-leaved species,tree density tended to decrease after each cutting. As a result, the stand type of these plots changed from a selection stand type to a single-layered stand type. In the conifer-dominated plot, it was predicted that the stand type could change from a selection stand type to a single-layered stand type, because of the increase in the broad-leaved tree density. These changes were mainly attributed to the suppression of tree regeneration induced by the thick growth of the Sasa senanensis dwarf bamboo. It was predicted that sustainable timber yields may not be secured in the future for any of the three plots. To continually conduct selection cutting management in stands where the forest floor is potentially covered by Sasa senanensis, it would be necessary to artificially promote tree regeneration in a cost-effective manner.