Thursday, August 5, 2010
Exhibit Hall A, David L Lawrence Convention Center
Masahiro Aiba, Takafumi Hino and Tsutomu Hiura, Tomakomai Research Forest, Hokkaido University, Tomakomai, Japan
Background/Question/Methods
Relative contribution of non-environmental factors, e.g. dispersal limitation and local extinction, to community assembly has been one of the major interests in community ecology in this decade. While many studies have tackled this problem in various plant communities, few studies have examined interspecific differences in determinants of spatial population structures among constituent species of a given community. Moderate contributions of both environmental and spatial factors to community structure, which have been repeatedly reported in many previous studies at community level, can arise from two different situations at population level. One is moderate contribution of both factors in most species and the other is a mixture of various species from mostly environmentally restricted to mostly spatially aggregated. If such interspecific difference exists, it is also interesting to relate the difference with functional, especially reproductive, traits of the species. To answer these questions, we examined relative contributions of environmental and spatial factors to community/population structure of understory plants in a semi-boreal forest using the latest, unbiased statistical procedures.
Results/Conclusions
In the community level analysis, unique contribution of environmental factors was 5.6%. Both contribution of spatial factors and shared variation between the environmental and spatial factors was 0%. Though the result suggests dominance of environmental factors as a determinant of the community structure at a glance, the result is partly attributable to the forward selection process before variation partitioning. Since effective spatial scale is largely different among species, most spatial variables which are important for a few species are omitted through the process. In population level analyses, spatial factors, which ranged from 0% to 21.1% (mean 4.6%), were also dominant components of explained variances for some species whereas environmental factors, which ranged from 0% to 21.4% (mean 6.6%), were dominant in more species. In 4 of the 24 analyzed species, the explained variances were purely spatial. These results clearly demonstrate that relative contributions of environment and space are considerably different among species. Therefore, the quest for links between determinants of spatial population structures and functional traits, e.g. life form, dispersal mode, seed size, ability of clonal reproduction, maximum height, specific leaf area and leaf nitrogen content will be an important future theme.