International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
symbiosis. The indices are described below, followed by the nu 
umerical criteria for evaluations in Table 1. 
Index 1: Vegetation Naturalness Index (VN): All plant 
communities within each unit were assigned a quantitative 
natural value. The average naturalness was then calculated 
by measuring the area within the unit occupied by each plant 
communities. 
N = Summation of [(Community Naturalness x Community 
Area)/ Small Watershed Area] 
The data is based on Natural Environment GIS (Ministry of 
Environment, 1998). The average naturalness was then 
calculated by measuring the area within the unit occupied by 
each plant community, according to the following formula. 
Index 2: Quantitative Index of Forest (QF): Quantitative index 
of the forest is defined as the proportion of forest area to the 
entire area of each small watershed. 
BF = Forest Area/Small Watershed Area 
Forest area is based on Natural Environment GIS (Ministry of 
Environment, 1998). 
Index 3: Index of Fragmentation of Natural Environment (FN) 
according to Road Effect 
Table 1. Evaluation of degree of natural symbiosis using four indices 
Roads were analyzed as a major factor in fragmentation and 
isolation of natural environments. The fragmentation index 
of the natural environment“ was calculated based on the total 
length of roads, defined as national motorways, national 
roads, prefectural roads and municipal roads included in the 
JMC Map (Japan Map Center, 1998). 
FN = Total Road Length/ Small Watershed Area 
Index 4: Land Development Rate (LD): To assess the 
magnitude of human impact on the natural environment, an 
index of land modification was calculated as the proportion 
of the total watershed areas occupied by artificial structures. 
LD = Land development Area/ Small Watershed Area 
The data regarding distribution of artificial structures is based 
on National Land Numerical Information (Natural Land 
Agency, 1997). 
The Total Score (Degree of Natural Symbiosis as calculated as 
VN+QF+FN+L D 
  
  
  
  
  
  
Evaluation 
Level VN QF LD Degree of Natural Symbiosis 
1 (7-10) (85-100) (0-0.05) (0-3) very high 
IT (3-7) (55-85) (0.05-0.1) (3-10) high 
I (1-3) (35-55) (0.1-0.15) (10-30) low 
IV (0-1) (0-35) (0.15- ) (30- ) very low 
  
  
  
  
  
  
  
  
3.3 Spatial Diversity Analysis 
The spatial diversity index was used to measure horizontal 
diversity between small watershed units. This technique, which 
uses measurements of Is (Interspersion) and J, (Juxtaposition) 
as components of spatial diversity, was described by Mead et al. 
(1981), and is considered to be the most effective index for 
quantitative and qualitative evaluation of habitats (Heinen & 
Cross, 1983). In this technique, calculations were originally 
implemented using raster (Clevenger et al, 1997, Clark et al., 
1993). In this research, however, measures of interspersion and 
juxtaposition were also used in a mesh analysis applied to 
polygon units within the small watershed units. In raster-based 
analysis, a value of 1 or 2 is respectively assigned to diagonal 
edges and to vertical or horizontal edges at juxtaposition by the 
raster. This can result in an underestimation. In addition, when 
the actual connectivity is considered, the length of a boundary 
between small watersheds can have a powerful influence on 
the suitability of the region as natural habitat, especially on 
large mammals such as black bears. It should also be kept in 
mind, however, that the connectivity calculated by either raster 
or grid does not have an equal influence on all species. Areas 
of utilization may vary among species. For these reasons, we 
also assessed spatial diversity utilizing polygons, which 
represent the actual situation as seen in the natural world. 
Masuyama et al. (2003) have shown that when the results of 
small watersheds evaluation are superimposed on a map 
showing the distribution of critical species, the areas with high 
degree of spatial diversity coincide with those in which these 
species were identified, proving the effectiveness of this 
evaluation method. Interspersion and Juxtaposition as spatial 
analyses provide an understanding of the connectivity among 
adjacent small watershed units. Following are descriptions of 
the original calculation method using grid, and consequently, 
small watersheds evaluation is suitable in conservation of 
biodiverse habitat. These spatial analyses provided an 
understanding of the connectivity among adjacent small 
watershed units. 
1) Original Calculation Method Using Grid 
1-1 Interspersion (I) 
L IL, HI and IV represent the small watershed evaluation 
categories as described in Table | above. Interspersion is 
calculated as the total number of changes recorded between 
adjacent units divided by the 
total possible number of 1 Il Il 
changes. In this case, for TL 
example, Is = 5/8 = 0.625 | 
I I I 
  
  
  
  
  
  
  
  
1-2 Juxtaposition (J,) 
Diagonal edges are assigned a score of 1; and either vertical or 
horizontal edges a score of 2. Various edge combinations can 
then be assigned a relative weight factors ranging from 0 to l. 
In the above case, for example, Jx = 6.6/12.0 = 0.55 
Edge types Quantity*' Quality*? Total 
/I 4 0.8 32 
Un 3 0.6 1.8 
IH 3 0.4 12 
[/IV 2 c2 0.2 104 
12 6.6 
2) Calculation Method Using Polygons p^ 
2-1 Interspersion (Isp) 
Isp is calculated as the number of different polygon evaluation 
levels divided by the total number of polygons. In this case, for 
example, 
260 
  
  
Internatior 
Titel tette 
Isp=3/4=0 
2-2 Juxtapt 
Quantitativ 
greater t 
perimeter/ [ 
Table 2. Ec 
Evaluation 
cells or poly 
4-1 Ecosys 
The resu 
symbiosis 
overall res 
watershed 
concentrate 
units with 
found alon 
the northez 
much lowe 
4-2 Spatial 
Analysis 
shows that 
high suitab 
some units 
are isolate 
indicate thz 
good conne 
Fig