Effects of disturbance intensity and frequency on bacterial community composition and function

Abstract

Disturbances influence community structure and ecosystem functioning. Bacteria are key players in ecosystems and it is therefore crucial to understand the effect of disturbances on bacterial communities and how they respond to them, both compositionally and functionally. The main aim of this study was to test the effect of differences in disturbance strength on bacterial communities. For this, we implemented two independent short-term experiments with dialysis bags containing natural bacterial communities, which were transplanted between ambient and 'disturbed' incubation tanks, manipulating either the intensity or the frequency of a salinity disturbance. We followed changes in community composition by terminal restriction fragment analysis (T-RFLP) and measured various community functions (bacterial production, carbon substrate utilization profiles and rates) directly after and after a short period of recovery under ambient conditions. Increases in disturbance strength resulted in gradually stronger changes in bacterial community composition and functions. In the disturbance intensity experiment, the sensitivity to the disturbance and the ability of recovery differed between different functions. In the disturbance frequency experiment, effects on the different functions were more consistent and recovery was not observed. Moreover, in case of the intensity experiment, there was also a time lag in the responses of community composition and functions, with functional responses being faster than compositional ones. To summarize, our study shows that disturbance strength has the potential to change the functional performance and composition of bacterial communities. It further highlights that the overall effects, rates of recovery and the degree of congruence in the response patterns of community composition and functioning along disturbance gradients depend on the type of function and the character of the disturbance.

Figure 1. NMDS plots showing the changes in BCC in relation to disturbance intensity (Experiment 1).

NMDS are based on Bray-Curtis distances according to bacterial community composition patterns (t-RFLP), after the disturbance (A) and at the end of the experiment (B). Error bars indicate standard errors.

Figure 2. NMDS plots showing the changes in BCC in relation to disturbance frequency (Experiment 2).

NMDS are based on Bray-Curtis distances according to bacterial community composition patterns (t-RFLP), after the disturbance (A) and at the end of the experiment (B). Error bars indicate standard errors.

Figure 3. Functional changes in the disturbance intensity experiment (Experiment 1).

Values presented correspond to mean values ± standard deviation of bacterial abundance (A, B), bacterial production (C, D), total carbon substrate utilization rate (tCSUR; E, F) and number of substrates with a positive response (G, H) for the different treatments. Mean values of controls are represented as dashed lines (- -), while standard deviation from control are represented as dotted lines (···). Tukey's homogenous groups are represented by a, b and c.

Figure 4. Functional changes in the disturbance frequency experiment (Experiment 2).

Values presented correspond to mean values ± standard deviation of bacterial abundance (A, B), bacterial production (C, D), total carbon substrate utilization rate (tCSUR; E, F) and number of substrates with a positive response (G, H) for the different treatments. Mean values of controls are represented as dashed lines (- -), while standard deviation from control are represented as dotted lines (···). Tukey's homogenous groups are represented by a, b and c.

Figure 5. Changes in the average slope of significant substrate guilds in the intensity experiment (Experiment 1).

Values plotted are the mean values ± standard deviation, after the disturbance (A–D) and at the end of the experiment (E–G). Mean values of controls are represented as dashed lines (- -) and standard deviations as dotted lines (···). Tukey's homogenous groups are represented by a, b and c.

Figure 6. Changes in the average slope of significant substrate guilds in the intensity experiment (Experiment 2).

Values plotted are the mean values ± standard deviation, after the disturbance (A–D) and at the end of the experiment (E–G). Mean values of controls are represented as dashed lines (- -) and standard deviations as dotted lines (···). Tukey's homogenous groups are represented by a, b and c.