The balance between deterministic and stochastic processes in structuring lake bacterioplankton community over time

Abstract

One major goal in microbial ecology is to establish the importance of deterministic and stochastic processes for community assembly. This is relevant to explain and predict how diversity changes at different temporal scales. However, understanding of the relative quantitative contribution of these processes and particularly of how they may change over time is limited. Here, we assessed the importance of deterministic and stochastic processes based on the analysis of the bacterial microbiome in one alpine oligotrophic and in one subalpine mesotrophic lake, which were sampled over two consecutive years at different time scales. We found that in both lakes, homogeneous selection (i.e., a deterministic process) was the main assembly process at the annual scale and explained 66.7% of the bacterial community turnover, despite differences in diversity and temporal variability patterns between ecosystems. However, in the alpine lake, homogenizing dispersal (i.e., a stochastic process) was the most important assembly process at the short-term (daily and weekly) sampling scale and explained 55% of the community turnover. Alpha diversity differed between lakes, and seasonal stability of the bacterial community was more evident in the oligotrophic lake than in the mesotrophic one. Our results demonstrate how important forces that govern temporal changes in bacterial communities act at different time scales. Overall, our study validates on a quantitative basis, the importance and dominance of deterministic processes in structuring bacterial communities in freshwater environments over long time scales.

Keywords: community assembly; 16S rRNA gene; bacterial community composition; mountain lakes; temporal dynamics.

Figure 1

(a) Asymptotic estimates of species richness, Shannon diversity, Simpson diversity and phylogenetic diversity (Faith's PD) for the bacterial community from Gossenköllesee (GKS) and Piburgersee (PIB). (b) changes in diversity estimates for GKS and PIB during the annual scale. Months in grey indicate the ice‐covered period [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 2

Bray–Curtis similarity of bacterial communities between November 2014 and December 2016 in Gossenköllesee (GKS) and between December 2014 and December 2016 in Piburgersee (PIB) (a), during the short‐term sampling in GKS (b) and for the water column (c) of both lakes. Seasonal trends (black lines) are shown. Months in grey indicate the ice‐covered period [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 3

Nonmetric Multidimensional Scaling (NMDS) analysis based on Bray–Curtis dissimilarity between the ice‐covered and ice‐free period in Gossenköllesee (GKS) and Piburgersee (PIB), showing the environmental parameters (significant parameters are shown with red arrows). Chl‐α, chlorophyll‐α. Cond, electrical conductivity. DN, dissolved nitrogen. DOC, dissolved organic carbon. O, dissolved oxygen. T, water temperature. TDP, total dissolved phosphorus [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 4

(a) Temporal changes in bacterial relative abundance in Gossenköllesee (GKS) and in Piburgersee (PIB). (b) The cladogram visualizes the output of the LEfSe algorithm, which identifies significant taxonomical differences between GKS and PIB [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 5

Contribution of different ecological processes to the assembly of the bacterial community in Gossenköllesee (GKS) and in Piburgersee (PIB) at the annual scale and at short‐term scale in GKS [Colour figure can be viewed at wileyonlinelibrary.com]