Gap formation and species diversity in Japanese beech forests: a test of the intermediate disturbance hypothesis on a geographic scale.

To evaluate whether the intermediate-disturbance hypothesis applies on regional scales, the relationship between the species diversity and gap formation regime of beech forests was examined. The mean gap size and the variation of gap sizes showed no correlation with species diversity. The mean windstorm interval varied widely, but geographical trends, such as latitudinal gradient, were not observed. However, locations that sustained an intermediate frequency of disturbance had the highest species diversity. Although a latitudinal gradient of disturbance was not apparent, the intermediate-disturbance hypothesis was partly supported on a geographic scale. The most predictable model for species diversity was a multiple regression model composed of two factors, the windstorm interval and the cumulative temperature of the growing season. The fact that the temperature was of greater importance than the disturbance interval indicates that the most important factor in predicting forest species diversity is the amount of available energy on a geographic scale.