Assessing niche separation among coexisting Limnohabitans strains through interactions with a competitor, viruses, and a bacterivore

Abstract

We investigated potential niche separation in two closely related (99.1% 16S rRNA gene sequence similarity) syntopic bacterial strains affiliated with the R-BT065 cluster, which represents a subgroup of the genus Limnohabitans. The two strains, designated B4 and D5, were isolated concurrently from a freshwater reservoir. Differences between the strains were examined through monitoring interactions with a bacterial competitor, Flectobacillus sp. (FL), and virus- and predator-induced mortality. Batch-type cocultures, designated B4+FL and D5+FL, were initiated with a similar biomass ratio among the strains. The proportion of each cell type present in the cocultures was monitored based on clear differences in cell sizes. Following exponential growth for 28 h, the cocultures were amended by the addition of two different concentrations of live or heat-inactivated viruses concentrated from the reservoir. Half of virus-amended treatments were inoculated immediately with an axenic flagellate predator, Poterioochromonas sp. The presence of the predator, of live viruses, and of competition between the strains significantly affected their population dynamics in the experimentally manipulated treatments. While strains B4 and FL appeared vulnerable to environmental viruses, strain D5 did not. Predator-induced mortality had the greatest impact on FL, followed by that on D5 and then B4. The virus-vulnerable B4 strain had smaller cells and lower biomass yield, but it was less subject to grazing. In contrast, the seemingly virus-resistant D5, with slightly larger grazing-vulnerable cells, was competitive with FL. Overall, our data suggest contrasting ecophysiological capabilities and partial niche separation in two coexisting Limnohabitans strains.

FIG. 1.

Microphotographs of the B4 and D5 strains (from the R-BT065 cluster of Betaproteobacteria) cocultured with FL (Flectobacillus sp. strain MWH38) documenting that the cells of the respective strains were clearly visually distinguishable in the cocultures. Typical representative morphotypes are highlighted in white frames. Scale bar, 5 μm.

FIG. 2.

Experimental design and major steps of manipulating the growth of studied bacterial strains by sources of bacterial mortality. At t₂₈, the treatments with B4+FL or D5+FL exponentially growing coculture were amended with heat-killed viruses (designated HV), with ∼3.5-times-more-concentrated heat-killed viruses (HVC), with live viruses (LV), and with ∼3.5-times-more-concentrated live viruses (LVC). In parallel, triplicates of all virus-amended treatments received a bacterivore (an axenic culture of Poterioochromonas sp.) capable of grazing on all cocultured bacterial strains (for details of cell size and morphology, see the legend to Fig. 1).

FIG. 3.

Time-course changes in bacterial abundance of the strains B4, D5, and FL in pure cultures (A) and in the cocultures B4+FL (B, D, F, and H) and D5+FL (C, E, G, and I) amended with heat-killed (B to E) or live viruses F to I) and a flagellate predator, Poterioochromonas sp. (D, E, H, and I). The vertical dashed line indicates experimental amendments at t₂₈. For a further explanation of the experimental design, see Fig. 2. Values are means from triplicate treatments, and vertical bars show ± 1 SD.

FIG. 4.

Time-course changes in bacterial biomass of the strains B4, D5, and FL in pure culture (A) and in the cocultures B4+FL (B, D, F, and H) and D5+FL (C, E, G, and I) amended with heat-killed (B to E) or live viruses (F to I) and a flagellate predator, Poterioochromonas sp. (D, E, H, and I). The vertical dashed line indicates experimental amendments at t₂₈. For a further explanation of the experimental design, see the legend to Fig. 1. Values are means from triplicate treatments, and vertical bars show ± 1 SD.

FIG. 5.

Time-course changes in abundance of B4 strains during the period after the sample manipulation (t₂₈ to t₈₅). (A) Example of the virus concentration effect, when the abundance of strain B4 was significantly lower at t₆₇ and t₈₅ (P < 0.01, Tukey's HSD test) in the LVC than in the basal live virus treatment (LV). (B) Trends in B4 abundance in HV and HVC treatments. (C) Example of the synergistic effect of both mortality sources (live viruses and a bacterivore) significantly accelerating mortality of B4 strains at t₆₇ and t₈₅ compared to the treatments amended by heat-killed viruses and a bacterivores (D). Treatments labeled + P, e.g., B4 − HV + P, received a bacterivore, Poterioochromonas sp. Values are means from triplicate treatments, and vertical bars show ± 1 SD.

FIG. 6.

Numerical ratio of the strains B4, D5, and FL at the end of the experiment (t₈₅) in the treatments with active viruses as related to their abundance in the treatments amended with heat-killed viruses in the absence (labeled no predation) or presence of a flagellate Poterioochromonas sp. (labeled + flagellate). However, in case of the FL strain grown in the presence of the flagellate, this ratio was calculated for the time point t₆₇, since the flagellate drastically decreased FL abundance (2 × 10⁴ to 6 × 10⁴ cells ml⁻¹) by the end of the experiment. The terms 1× and 3.5× refer to basal and concentrated virus additions, respectively. Values are means from triplicate treatments, and vertical bars show ± 1 SD.

FIG. 7.

Biomass yield of the strains B4, D5, and FL in cocultures amended with bacterial mortality factors expressed as a percentage of the maximum biomass yield achieved in a monoculture (see the control in Fig. 4A) of the respective strain grown in the same medium (40 mg of NSY medium liter⁻¹). The absence (labeled no predation) or presence of a flagellate Poterioochromonas sp. (labeled + flagellate) is indicated. Since no significant effect of two different virus concentrations on maximum biomass yield was detected (P > 0.05, t test), only data for the basal virus concentration are shown. Values are means from triplicate treatments, and vertical bars show ± 1 SD.