5) Organization of biological communities: test of the modern unified theory
5.1 Checking reliability of statistical methods used in community assembly studies by analysis of simulated data
In previous work (Botta-Dukát & Czúcz 2015), we have shown that some of the widely used statistical method are not reliable: their Type I or Type II error rate is too high. The main message of that work is that all methods should be checked using simulated data. In the recent project we will continue this work by testing further methods.
The majority of our current knowledge on the organization of grassland communities are based on plant species and vegetation. In this study we use data on arthropod and vertebrate animal species from a large, landscape-scale habitat restoration and management program (3.2.) in a unified framework and analysis to test those predictions of the most important theories of community organization that bear on taxonomic and functional diversity as well as species composition.
Trait-based methods allow estimating the relative role of local (environmental filtering, limiting similarity) and landscape level (mass effect) factors in the assembly of communities and detecting relationship between traits and processes. In a previous project such methods were successfully applied in the analysis of terrestrial plant communities. In the recent project we will apply them for other type of communities for searching general trends.
Impact of the lake size and habitat diversity on the diversity of planktonic and benthic algae is studied. The importance of adjacent habitats in maintaining macroinvertebrate diversity of permanent freshwater marshy ecosystems is being investigated. Relative roles of environmental and spatial factors in determining macroinvertebrate metacommunity structure in riverine systems is being modelling.
We study the main drivers determining the composition and diversity of the benthic diatom
communities in small astatic ponds.