as character species of Fagetea crenatae, and 
Camellia japonica (Japanese camellia) and 
Castanopsis cuspidata var. Sieboldii 
(Japanese chinquapin) were analyzed as 
character species of Camellietea japonicae. 
Firstly, relevés where each species 
occurs were extracted, which consist of 
relevés of beech (46 second grids), oak (62 
grids), chinquapin (34 grids), and camellia 
(66 grids). 
Then, the map names and positions on 
the map were extracted from the relevé 
sheets mentioned above, and they were 
converted into the second grid code by 
means of the Grid-Cell Code Data File. The 
distribution of the plant species was mapped 
by using the second grid-cell code. Figure 4 
shows the points where beech or oak are 
distributed. They are distributed mainly in 
the northern part of the Japanese 
archipelago; however, they are also 
distributed in the mountainous region of 
Shikoku and Kyushu. Figure 5 shows the 
points where chinquapin or camellia is 
distributed. They are distributed mainly in 
the southern part of the Japanese 
archipelago; however, they also reach 
northern Honshu near the coastal zone. 
Next, the climatic data, which 
corresponds to the second grid-cell codes, 
were extracted from the Natural Environment 
Data File. Since this Natural Environment 
Data File uses the third grid as a unit, each 
second grid code has 100 observations. 
Here, the mean values were calculated from 
100 observations. Figure 6 takes the mean 
annual temperature as the horizontal axis, 
and the percentage of the occurrence of each 
plant species as the vertical axis. From 
Figure 6, the habitat segregation between 
beech-oak and chinquapin-camellia by the 
temperature factor can be found, and the 
boundary is approximately 12 - 14 degrees 
Centigrade. Similarly, Figure 7 takes the 
annual precipitation as the horizontal axis. 
The annual precipitation does not appear to 
affect the distribution. 
The above-mentioned analysis results 
are not contradictory to the ecological 
findings. However, because other habitats 
are reported than the places shown in Figure 
4 and Figure 5, it is necessary to increase 
amount of data further and to improve the 
accuracy of the analysis. 
  
Data Entry 
  
Personal-Computer 
* Digitizing of relevé 
(Japanese word processor) 
Data Check ( C program) 
* List up and Sort of species names 
Data Analysis ( C program) 
* Pick up relevé by species names 
and List up map names 
* Convert map name to second grid code 
* Mapping of distribution 
  
  
  
Terra-Mar 
Image Processing 
* Map Display 
* Three dimensional display 
* Zoom 
* Overlay 
* Clasification 
  
Main Frame Computer 
Statistics (SAS) 
* Basic Statistics 
* Correlation 
* Factor Analysis 
* Discriminant Analysis 
Modeling (Fortran program) 
* Fuzzy Model 
* Multinominal Logit Model 
  
  
  
  
  
  
Figure3 Composition of Software 
54