Search

Additional file 3. Supplemental discussion and methods

Additional file 1: Supplementary figure 1: Sampling locations for all SPRUCE samples. Sampling locations within the S1 Bog at the Marcell Experimental Forest in Northern Minnesota, USA, including the five transect samples and the samples from the SPRUCE experimental chambers. Numbers next to the brackets show how many and what kinds of metagenomes were derived from each part of the bog. Supplementary figure 2: A: Network of shared predicted protein content between recovered SPRUCE viruses (n = 4,326), and RefSeq prokaryotic viral genomes (n = 37). Colored nodes represent vOTUs, nodes are colored by the dataset(s) from which they were recovered, and the shared edges represent shared predicted protein content. B: Distribution of vOTUs into VCs, recovered from each of the three extraction methods and collection dates. Numbers represent number of VCs that contain vOTUs from the extraction method(s) listed. Supplementary Figure 3: Taxonomic classification of soil vOTUs in PIGEON. Taxonomic classifications were based on vConTACT2.0 clustering with RefSeq prokaryotic viral genomes. Percentages at the top of each graph indicate the proportion of vOTUs that were taxonomically classified, n represents the total number of vOTUs that could be taxonomically classified. Supplementary figure 4: Comparison of vOTU recovery from five paired viromes and total soil metagenomes from the SPRUCE transect. A: Distribution of vOTUs recovered by each of the two extraction methods, based on read mapping to the PIGEON database, including all vOTUs recovered from SPRUCE. B: Accumulation curves of distinct vOTUs recovered as sampling increases for each extraction method; 100 permutations of sample order are depicted as open circles, and averages are shown as a line. C: Similar to panel B, but only the accumulation curve of distinct vOTUs recovered from total soil metagenomes is shown, with a smaller y-axis maximum to better show the trend. Supplementary figure 5: Comparison of the five viromes from the transect. A: Dendrogram depicting sample similarity according to viral community composition (left) and heatmap (right) of vOTUs detected (green = detected, white = not detected) in the five SPRUCE transect viromes. B:Comparison of vOTU recovery from the SPRUCE-2 sample compared to the four other virome samples.

In disturbance ecology, stability is composed of resistance to change and resilience towards recovery after the disturbance subsides. Two key microbial mechanisms that can support microbiome stability include dormancy and dispersal. Specifically, microbial populations that are sensitive to disturbance can be re-seeded by local dormant pools of viable and reactivated cells, or by immigrants dispersed from regional metacommunities. However, it is difficult to quantify the contributions of these mechanisms to stability without, first, distinguishing the active from inactive membership, and, second, distinguishing the populations recovered by local resuscitation from those recovered by dispersed immigrants. Here, we investigate the contributions of dormancy dynamics (activation and inactivation), and dispersal to soil microbial community resistance and resilience. We designed a replicated, 45-week time-series experiment to quantify the responses of the active soil microbial community to a thermal press disturbance, including unwarmed control mesocosms, disturbed mesocosms without dispersal, and disturbed mesocosms with dispersal after the release of the stressor. Communities changed in structure within one week of warming. Though the disturbed mesocosms did not fully recover within 29 weeks, resuscitation of thermotolerant taxa was key for community transition during the press, and both resuscitation of opportunistic taxa and immigration contributed to community resilience. Also, mesocosms with dispersal were more resilient than mesocosms without. This work advances the mechanistic understanding of how microbiomes respond to disturbances in their environment. This article is part of the theme issue ‘Conceptual challenges in microbial community ecology’.

Press disturbances are stressors that are extended or ongoing relative to the generation times of community members, and, due to their longevity, have the potential to alter communities beyond the possibility of recovery. They also provide key opportunities to investigate ecological resilience and to probe biological limits in the face of prolonged stressors. The underground coal mine fire in Centralia, Pennsylvania has been burning since 1962 and severely alters the overlying surface soils by elevating temperatures and depositing coal combustion pollutants. As the fire burns along the coal seams to disturb new soils, previously disturbed soils return to ambient temperatures, resulting in a chronosequence of fire impact. We used 16S rRNA gene sequencing to investigate bacterial and archaeal soil community responses along two active fire fronts in Centralia, and investigated the influences of assembly processes (selection, dispersal and drift) on community outcomes. The hottest soils harbored the most variable and divergent communities, despite their reduced diversity. Recovered soils converged toward similar community structures, demonstrating resilience within 10-20 years and exhibiting near-complete return to reference communities. Measured soil properties (selection), local dispersal, and neutral community assembly models could not explain the divergences of communities observed at temperature extremes. We hypothesize that transitions between the seed bank and the active community, which would manifest as drift processes, are key in explaining these divergences. These results suggest that soils generally have an intrinsic capacity for robustness to varied disturbances, even to press disturbances considered to be “extreme”, compounded, or incongruent with natural conditions.

In disturbance ecology, stability is composed of resistance to change and resilience towards recovery after the disturbance subsides. Two key microbial mechanisms that can support microbiome stability include dormancy and dispersal. Specifically, microbial populations that are sensitive to disturbance can be re-seeded by local dormant pools of viable and reactivated cells, or by immigrants dispersed from regional metacommunities. However, it is difficult to quantify the contributions of these mechanisms to stability without, first, distinguishing the active from inactive membership, and, second, distinguishing the populations recovered by local resuscitation from those recovered by dispersed immigrants. Here, we investigate the contributions of dormancy dynamics (activation and inactivation), and dispersal to soil microbial community resistance and resilience. We designed a replicated, 45-week time-series experiment to quantify the responses of the active soil microbial community to a thermal press disturbance, including unwarmed control mesocosms, disturbed mesocosms without dispersal, and disturbed mesocosms with dispersal after the release of the stressor. Communities changed in structure within one week of warming. Though the disturbed mesocosms did not fully recover within 29 weeks, resuscitation of thermotolerant taxa was key for community transition during the press, and both resuscitation of opportunistic taxa and immigration contributed to community resilience. Also, mesocosms with dispersal were more resilient than mesocosms without. This work advances the mechanistic understanding of how microbiomes respond to disturbances in their environment.This article is part of the theme issue ‘Conceptual challenges in microbial community ecology’.

In disturbance ecology, stability is composed of resistance to change and resilience towards recovery after the disturbance subsides. Two key microbial mechanisms that can support microbiome stability include dormancy and dispersal. Specifically, microbial populations that are sensitive to disturbance can be re-seeded by local dormant pools of viable and reactivated cells, or by immigrants dispersed from regional metacommunities. However, it is difficult to quantify the contributions of these mechanisms to stability without, first, distinguishing the active from inactive membership, and, second, distinguishing the populations recovered by local resuscitation from those recovered by dispersed immigrants. Here, we investigate the contributions of dormancy dynamics (activation and inactivation), and dispersal to soil microbial community resistance and resilience. We designed a replicated, 45-week time-series experiment to quantify the responses of the active soil microbial community to a thermal press disturbance, including unwarmed control mesocosms, disturbed mesocosms without dispersal, and disturbed mesocosms with dispersal after the release of the stressor. Communities changed in structure within one week of warming. Though the disturbed mesocosms did not fully recover within 29 weeks, resuscitation of thermotolerant taxa was key for community transition during the press, and both resuscitation of opportunistic taxa and immigration contributed to community resilience. Also, mesocosms with dispersal were more resilient than mesocosms without. This work advances the mechanistic understanding of how microbiomes respond to disturbances in their environment.This article is part of the theme issue ‘Conceptual challenges in microbial community ecology’.