267 
MODELING OF SPECIES GEOGRAPHIC DISTRIBUTION 
FOR ASSESSING PRESENT NEEDS FOR THE ECOLOGICAL NETWORKS 
T. Doko a ’ *,F A. Kooiman\ A.G.Toxopeus b 
a Graduate School of Media and Governance, Keio University, 
302 Z-building of Keio University, 5322 Endoh, Fujisawa, Kanagawa, 252-8520, Japan - dokochan@sfc.keio.ac.jp 
b Department of Natural Resources, International Institute for Geo-information Science and Earth Observation (ITC), 
P. O. Box 6, Hengelosestraat 99, 7500 AA, Enschede, The Netherlands - kooiman@itc.nl, toxopeus@itc.nl 
KEY WORDS: Ecology, Environment, GIS, Modelling, Algorithms, Landscape, Method, Proposal 
ABSTRACT: 
In Japan, attention is currently focused on designing ecological networks for wildlife animals. However there is an obvious lack of 
the species spatial information. This study aims (a) to acquire the potential spatial distribution of Asiatic black bear and Japanese se- 
row to identify core areas, and (b) to propose a methodology for assessing needs for ecological networks. 1836 species’ point records 
and 14 potential predictors were prepared in a GIS environment, split into a train and a test dataset. Screening predictors by statisti 
cal analysis, we modeled species geographic distribution by three algorithms: GARP, MaxEnt, and GLMs in Kanagawa and Shizu 
oka Prefectures. Based on the most accurate maps, assessed by Kappa statistics, population was estimated based on population den 
sity and area of habitat patch. For bear, MaxEnt performed best with the predictor variables: altitude, distance to paths and stone 
steps, distance to wide roads, and vegetation cover types. GARP failed to predict presence in Fuji. Its best GLM equation was 
log(p/(l-p))=(-1.486e+01)+(7.335e-04)*<7Aia«ce to paths and stone steps+(9A10e-03)*altitude. For serow’s distribution, GARP per 
formed best with altitude, slope, distance to highways, distance to general roads, distance to paths and stone steps, distance to rivers, 
and NDVI. Its best GLM equation was log(p/( 1 -/?))=-5.91785430+0.04024 \36*slope+0.26478159*square root of altitude. The esti 
mated numbers of individuals for bear was 5-9 in Mt. Ashitaka, 51 ~ 102 in Fuji-Tanzawa, 160-320 in South Alps, 4-8 in Mt. Ke- 
nashi, 4-8 in Izu Peninsula, and 6-11 in Hakone; for serow, <1581 were estimated in Fuji-Tanzawa, and < 537 in other areas. For 
bear MaxEnt and for serow GARP are the best algorithms, but GLM has good transferability. There is a need for ecological net 
works in Fuji-Tanzawa for bear, but not for serow. 
1. INTRODUCTION 
1.1 Biodiversity loss and ecological networks 
The New Biodiversity Strategy of Japan (March 27, 2002) 
stipulates seven major themes for implementing biodiversity 
conservation policies. The first theme is conservation of prior 
ity areas and formation of ecological networks. Its basis is to re 
inforce the protected-area system. In addition to the perspective 
of conserving natural landscape of the Natural Parks, measures 
from the perspective of ecosystem conservation, especially of 
animal habitat conservation, should be institutionalized. The 
development of the ecological networks can connect the frag 
mented habitats of wild animals, stem the biodiversity loss, and 
promote dispersal and genetic exchange of wild species. Seri 
ous fragmentation of habitats has been caused by the industri 
alization of agriculture, restructuring of land use, the building 
of huge transport networks and metropolitan (Stanners and 
Bourdeau, 1995). For fragmented habitats, the theory of island 
biogeography (MacArthur and Wilson, 2001) can be applied, 
and connecting the ‘islands’ through the ecological networks 
can reduce the risk of extinction of species. 
As for the Japanese mammals, the Ministry of Environment of 
Japan has conducted the national distributional survey of Japa 
nese animals (Biodiversity center of Japan, 2004) in 1978 and 
in 2003 as a monitoring activity. The main objective was to ac 
quire national distributional maps of ten main mammals, in 
cluding key wildlife species in this study. Based on survey data 
from interviews and questionnaires on a sampling grid of 5 by 5 
km, these distribution maps were compiled at a national scale 
(1: 2,500,000). 
Although these maps provide insight in species’ distribution at 
a glance, they do not reflect distribution at local population 
level. An appropriate approach for preparing conservation and 
zoning plans requires spatially explicit information of species 
distribution at a local scale with more accurate resolution. With 
this information, the core area of a suitable habitat can be iden 
tified, and ecological networks can be designed if necessary. 
There are several studies that have integrated habitat models 
into GIS (for example, see Corsi et al., 1999). However, these 
have not assessed the need for ecological networks of large 
mammals. Also, to date, no studies were found on quantitative 
needs assessment of ecological networks. 
In this paper, we present a quantitative methodology for the 
modeling of the geographic distribution of two key species of 
large mammals in order to assess the need for ecological net 
works. 
1.2 Research objectives 
There are three main objectives of this study: 1) to acquire 
more accurate potential spatial distribution of habitats of ’key’ 
wildlife species, 2) to identify the ‘core areas’ of these habitats, 
and 3) to assess the present needs for ecological networks. 
Corresponding author.