Environmental filtering drives assembly of diatom communities over evolutionary time‐scales

Aim Ecological communities are structured through the interplay of deterministic assembly processes such as competition and environmental filtering. Whereas the drivers of spatial community structure are frequently studied in extant communities, little is known about the relative importance of assembly processes in response to environmental factors over evolutionary time-scales. Here, we use an integrative framework to unravel community assembly processes since the inception of a long-lived lake ecosystem. Location Lake Ohrid. Time period From lake formation 1.36 million years ago to the present. Major taxa studied Planktonic diatoms. Methods We constructed a dated phylogeny of extant and extinct diatoms and collected trait data for 380 fossil communities to quantify phylogenetic community structure and functional richness and to determine the relative importance of deterministic assembly processes over time. We then used regression analysis to correlate the phylogenetic community structure with palaeoenvironmental and intrinsic biological predictors and to identify primary drivers of assembly processes. Results Our results suggest a dense packing of niche space with higher species richness and co-occurrence of closely related species. There are only two short episodes in the very recent past dominated by distantly related taxa. We found distinct changes in phylogenetic community structure upon speciation or extinction events and an increase in mean community relatedness over time. Main conclusions Our finding of closely related co-occurring species implies environmental filtering as the primary assembly mechanism, with a minor but increasingly important role of competition towards the present, driven by evolutionary dynamics. Such an increase in the relative contribution of competition to the assembly of communities in relation to the aging of an insular ecosystem, together with a denser packing of morphospace in the early phase of system ontogeny is compatible with ecological predictions according to the theory of island biogeography.