Consistent Patterns of Statistical Distributions in Natural Ecological Communities: Lake Phytoplankton

Abstract

In spite of the very high time-space heterogeneity of lacustrine phytoplankton, a high level of orderliness of the phytoplankton taxonomic structure was found with the help of several approximation models. Stochastic dynamics of abundances and biomasses often emerges from mathematical models and experiments considering a small number of interacting species. Conversely, some predictable patterns and indices describing aquatic assemblages emerge from large-scale studies. Models with an extremely small number of parameters are discussed, applicable as good approximations for the annual and multi-annual taxonomic-unit distributions of lacustrine phytoplankton. Several statistical models (i.e., rank-abundance, rank-biomass, rank-frequency, and rank-size distributions) of the whole assemblage were built using 8-year monitoring data of Lake Kinneret phytoplankton. Due to the high coefficient of determination of linear regression, long-term geometric-series rank distribution models were selected as the best. Lognormal distributions produced by the scale-free food web lead to an explanatory model based on concepts of the living whirl (G. Cuvier), living matter (V. Vernadsky), and dissipative structures.