Positive effects of forest fragmentation per se on bryophyte diversity in subtropical fragmented forests: evidence from land-bridge islands

Yuan Tonghe Dandan Li Shan Lv Heng Zhang Yuzhu Xia Xiao Wang Jun Yang Jing Yu ^* Shuiliang Guo ^*

• Shanghai Normal University, Shanghai, China

Habitat fragmentation (Sensu lato) represents a landscape-scale process involving both habitat loss and the breaking apart of habitat (habitat fragmentation per se). To explore the effects of habitat fragmentation per se on bryophyte assemblages in subtropical forests, we investigated bryophytes and environments on 18 fragmented forest landscapes (including 166 islands) in Thousand Island Lake, China. Landscape-level environmental variables of habitat fragmentation per se included island number, mean area, area variability, shape irregularity, shape variability, and isolation degree. Landscape-level habitat amount was represented by island total area within the study landscape. We investigated species richness (SR) and coverage in edge zones and interior environments of thirteen islands to explore the edge effects of fragmented forests on bryophytes. Variance partitioning revealed that habitat fragmentation per se independently explained 38.92% of variation in bryophyte SR and 36.5% of variation in species composition (SC). Landscape-level island total area explained 6.2% of SR variation and 5.9% of SC variation. Among the environmental variables associated with fragmentation per se, island number and shape irregularity were identified as the most significant, independently explaining 16.2% and 15.5% of variation in bryophyte SR, respectively. Island shape variability and area variability independently explained 5.3% and 2.1% variations in bryophyte SR, respectively. A linear increase in bryophyte SR was observed with island mean area and shape irregularity, while a nonlinear relationship was detected with island number, island shape irregularity and area variability. Island area variability, shape variability and island number influenced bryophyte SC to similar extents, independently explaining 5.9% to 6.6% of variation in bryophyte SC. Consequently, habitat fragmentation per se had pronounced effects on both bryophyte SR and SC in subtropical fragmented forests. Such effects were likely due to the positive edge effects of fragmented forests on bryophyte assemblages. Our findings suggest that, in subtropical fragmented secondary forests, the reserve for bryophytes had better contain numerous forest patches with irregular shapes, large total area, and moderate variation in island shape and area.