Biotic homogenization has been reported worldwide. Although simplification of communities across space is often significant at larger scales, it could also occur at the local scale by changing biotic interactions. This study aimed to elucidate local community processes driving biotic homogenization of soil faunal communities, and the possibility of biotic re-differentiation. We recorded species of oribatid mites in litter and soil layers along a gradient of forest conversion from monoculture larch plantation to primary forests in central Japan. We collected data for functional traits of the recorded species to quantify functional diversity. Then we quantified their taxonomic/functional turnover. Litter diversity was reduced in the larch-dominated stands, leading to habitat homogenization. Consequently, litter communities were biologically homogenized and differentiated in the plantations and in the natural forest, respectively. Turnover of functional traits for litter communities was lower and higher than expected by chance in the plantations and in the natural stand, respectively. This result suggests that the dominant assembly process shifts from limiting similarity to habitat filtering along the forest restoration gradient. However, support for such niche-based explanations was not observed for communities in the soil layer. In the monocultures, functional diversity expected from a given regional species pool significantly decreased for litter communities but not for those in the soil layer. Such discrepancy between communities in different layers suggests that communities more exposed to anthropogenic stresses are more vulnerable to the loss of their functional roles. Our study explains possible community processes behind the observed patterns of biological organization, which can be potentially useful in guiding approaches for restoring biodiversity.

We reconstructed the vegetational landscape of the pre-industrial era (the beginning of the twentieth century) in north-eastern Japan, and estimated the distribution patterns of traditional land-uses, as suggested from the vegetation. We found significant correspondence between the spatial patterns of vegetational landscape and site attributes, and hypothesized the underlying mechanisms. The study area was classified into three vegetation types: grasslands, secondary forests and old-growth forests. It was determined that the grasslands were formed and maintained by burning; secondary forests were derived from either charcoal woods or forests recovered on abandoned grasslands; and oldgrowth forests had suffered the least anthropogenic disturbance. Each past vegetation type showed significant dependency on site attributes such as altitude, slope angle, slope aspect, hydrological topography and distance from the nearest human habitation. The relative importance of these site attributes varied depending on the vegetation type. Grasslands and old-growth forests, which were the most and the least disturbed sites in the study area, respectively, showed clear contrasts in their dependencies especially on the slope aspect and on elevation. These site attributes were thought to have had influences on each vegetation type by determining the inflammability of the site. Satellite photographs indicated that north-facing valleys had been relatively wet throughout the fire-prone spring season. Hence, these areas would have been free from frequent fire, and more likely to preserve old-growth forests. Ground wetness in spring was thought to be the underlying factor determining the contrasts in past vegetation and land-use patterns in the area.

No abstract

No abstract

No abstract