Geographic Assessment of Present Protected Areas in Japan for Representativeness of Forest Communities

Most assessments of present protected areas have focused on which features are or are not represented and to what extent, but they have not considered the environmental gradients and the geographic context within each biodiversity feature under a conservation network. We examined how protected areas are distributed with respect to the distribution of six forest community types. Three analyses were applied to the dataset in the Red Data Book of Plant Communities in Japan and the related survey: (1) recursive partitioning was used to contrast environmental factors of conserved communities with nonconserved communities; (2) point pattern analysis, based on Ripley's K function, was used to describe the spatial pattern of conserved communities; and (3) the spatial scan statistic was used to detect spatial representation gaps. Overall, environmental bias was greatest in relation to soil and topography. However, the results of point pattern analysis showed that the spatial pattern of conserved communities did not depend entirely on the distribution of environmental factors. Four types of gaps in spatial representation were detected by the spatial scan statistic, irrespective of environmental bias. These results showed that although a community type might be well protected in total or along the environmental gradients, conserved communities might not capture the full range of geographic context. To ensure appropriate representation or protection, it is important that conservation planning for protected areas take into account both the environmental gradients and the geographic context within each biodiversity feature.