A whole ecosystem manipulation approach to study the effect of disturbances on bacterial metacommunities

Background/Question/Methods

Effects of disturbances on biodiversity and ecosystem functioning have primarily been studied at the local scale, whereas much less is known about effects at the metacommunity scale, in particular for bacterial communities in natural systems. Therefore we implemented a whole ecosystem manipulation experiment with rock pools as natural model systems to test how a salinity disturbance that was applied at different strengths influences the taxonomic and functional diversity of bacterial communities and their resistance and resilience. Moreover, this experiment aimed to disentangle the influence of dispersal and colonisation by air- and rainborne bacteria for the recovery of the local communities. Freshwater rock pools were chosen and their salinity was manipulated by adding salt at four different levels with 6 pools per salinity level: no increase, 3 ppm 6 ppm and 12 ppm. Pools were covered for 62 hours and subsequently opened again, and then allowed to re-fill with rain. Samples were taken before the salt addition, directly after the cover was removed and, finally, when pools had reached their initial salinities again (after 96 hours). Community respiration, functional diversity (measured as substrate utilisation profiles using Biolog Ecoplates) and community composition (454 sequencing of 16S rRNA) were measured.

Results/Conclusions

Increasing salinities caused changes at the functional level that were strongest in the treatment with the highest salinity and tended to recover at low and intermediate levels. However, these patterns were not accompanied by changes at the community level, since both average local taxa richness and beta-diversity remained unaltered. Moreover, bacterial communities in air-and rain samples were clearly different from those in the pools indicating that they had no major effect on modifying bacterial diversity during the re-filling period.  Our results indicate that the applied salinity disturbance affected only community functioning but not composition and did, hence, not result in taxa losses or directional and synchronous changes in bacterial community composition among freshwater rock pools. This result might reflect the fact that salinity fluctuations occur naturally in rock pools, which leads to compositional resistance and functional resilience of their bacterial communities to salinity perturbations.