Recently, the distribution of sika deer (
Cervus nippon) has been expanding, and the consequent damages to crops and forests are increasing. Clarification of factors influencing this expansion might help in controlling an expansion of damaged areas. In this analysis, we hypothesized that the variables of expanding distribution vary according to the distance from source habitats. Therefore, we attempted to reveal the factors of expanding distribution by using several analysis ranges.
A generalized linear model (GLM) was used to derive predictive habitat distribution models from vegetational, topographical, and social predictors in a geographic information system (GIS), using the sika deer data in Honshu, Shikoku, and Kyushu, Japan, collected in the national survey on natural environment in 1978 and 2003.
The results showed that snow depth and population were important variables in the whole area model without distance from source habitats. On the other hand, distance was strongly associated with the distribution in the whole area model with distance. Furthermore, results from models employing smaller analysis ranges revealed that the sika deer distribution is influenced by cultivation abandonment, conifer and roads. This result implies that to prevent an expansion of damaged areas, cultivation abandonment and conifer should be managed.
View full abstract