| Summary: | Abstract Knowing which factors govern variation in community structure is crucial for assessing and conserving biodiversity. Two main processes structuring biological communities are selection by environmental conditions and dispersal between sites. In this thesis, I studied the potential importance of these two processes in structuring stream metacommunities (i.e. a set of local communities that are connected by the dispersal of individuals). I used altogether four datasets collected from three different drainage basins located in northern Finland. For small species, such as stream diatoms, bryophytes and macroinvertebrates, measuring dispersal directly and for all the species in a community, is very difficult. Thus, I used three kinds of spatial proxies to represent the potential effects of dispersal: (1) the size of species pool, (2) spatial locations of the sample sites and (3) distances between sites. In general, local environmental characteristics explained the variation in community structure better than any of the spatial proxies. The results suggested that stream diatoms, bryophytes and macroinvertebrates should not be dispersal limited at within-basin scale. However, in some cases spatial proxies explained a relatively larger proportion of community variation. One major difficulty when using such proxies is that high and limiting dispersal rates produce similar patterns. Here, I also present and test a novel hierarchical metacommunity approach for differentiating if a significant spatial signal is produced by high or limiting dispersal rates. This approach seemed promising. The results suggested that, at very small spatial scales (i.e. within a stream section) the community composition of stream macroinvertebrates is possibly influenced by high dispersal rates. I also used the so-called deconstructive approach and assigned the whole community into smaller groups of species based on their traits (i.e. feeding mode, dispersal mode and body size). This approach showed, for example, that metacommunity ...
|
|---|