STATISTICAL APPROACHES FOR SPATIAL-TEMPORAL DYNAMICS OF MICROBIAL COMMUNITIES IN THE RED SEA

Abstract

Changing environmental factors impact the biological structure and function of marine microbial communities. There is a need for modelling samples of microbial communities as a mixture of sub-communities. Sub-communities are modelled to be consistent with the observed distribution and abundance of taxa because (i) there is extensive among-sample variation in microbial abundances, and (ii) there are important ecosystem processes that appear to be carried out by the collective activities of microbial species. Three statistical frameworks are applied to identify a robust consensus of sub-community structure from the Gulf of Aqaba, Red Sea. Assemblages of taxa are derived from different methods to capture sub-community spatial-temporal dynamics and can be used to train predictive models of important environmental categories. Assemblages and their test predictions are compared to each other and to results from variable selection of individual microbial taxa. Assemblages also demonstrate characteristics associated with distinct physiochemical features of the ecosystem.