Abstract
To establish an accurate method for estimating the population density of the Japanese giant flying squirrel Petaurista leucogenys, we estimated the population density of an isolated population in Machida, Tokyo by two methods : the line transect census and the use ratio of tree hollow nests. The density obtained from the line transect census was 0.41/ha at evergreen homestead woodlands and 0.03/ha at deciduous Satoyama, averaging 0.08/ha for the total area. The distribution of tree hollow nests was much more dense at homestead woodlands and the nest use ratio averaged 47%. The estimated density from the nest use ratio was 2.0/ha at homestead woodlands and 0.22/ha at deciduous Satoyama. Although the reason for the discrepancy between the two estimation methods is uncertain, both clarified that survey according to vegetation types was indispensable. At least three surveys were necessary to obtain reliable results by the line transect census. The line transect census seemed more suitable in terms of cost efficiency and applicability to diverse forest types.
