BERIL Knowledge Wiki

Concept in Subsurface and Soil Community Ecology

Ecological memory preserves community differences across changing environments

Single-source. AI-generated wiki from AI-conducted research reports. Partially reviewed by scientists. See About This Wiki.

Projects cited
1 project report, listed in the margin
Linked from
7 pages5 topics

Ecological memory is the persistence of community-composition differences caused by prior environmental exposure after communities encounter the same current conditions. The lignin-enrichment experiment provides a quantitative bacterial example: Round-1 carbon history explained more Round-2 community variation than the current Round-2 carbon source, and communities with different histories did not converge under identical conditions. 1

Core Evidence

The study analyzed 21 samples across 7 groups, with 3 replicates per group, using bacterial 16S V3–V4 and fungal ITS2 (internal transcribed spacer) amplicon data, 97% OTU (operational taxonomic unit, a sequence cluster used as a taxon proxy) clustering, Bray–Curtis distances, PCoA (principal coordinates analysis, an ordination of the distance matrix), and PERMANOVA, a permutation-based test of community-composition differences. 1 The Round-2 factorial PERMANOVA of bacterial 16S communities included 12 samples and found that Round-1 history explained 58.9% of community variance (F=14.31, R²=0.589, p=0.002), whereas the current Round-2 carbon source explained 32.7% (F=4.85, R²=0.327, p=0.018). 1 Communities with different Round-1 histories did not converge even under identical Round-2 conditions, and the history/Round-2 distance ratio was 1.15–1.59, where values >1 indicate that history matters more than current conditions. The report presents this as quantitative evidence for ecological memory. 1 This quantification applies to bacteria only: in fungal ITS communities the Round-1 history signal was not statistically detectable (p=0.090), so the result does not establish ecological memory in fungi. 1

Across the three-tier PERMANOVA design (999 permutations, Bray–Curtis), the all-7-group model (M3) gave 16S F=111.35, R²=0.979, p=0.001 and ITS F=3.02, R²=0.582, p=0.001. The Base-versus-Round-1 model (M1) gave 16S F=394.42, R²=0.992, p=0.008 and ITS F=6.92, R²=0.698, p=0.003. The Round-2 factorial group term (M2) gave 16S F=41.12, R²=0.939, p=0.001 and ITS F=1.72, R²=0.424, p=0.011. 1 Within the Round-2 factorial, neither ITS factor was detected: the ITS Round-1 history term gave F=1.49, R²=0.142, p=0.090 and the ITS Round-2 labile-carbon term gave F=1.71, R²=0.160, p=0.090. These null results contrast with the significant 16S history (F=14.31, R²=0.589, p=0.002) and labile-carbon (F=4.85, R²=0.327, p=0.018) terms. 1

Kruskal–Wallis tests showed that alpha diversity differed across all 7 groups for all metrics (16S: p=0.006–0.010; ITS: p=0.015–0.041). 1 PCoA ordination showed clear separation of all treatment groups, with PC1 and PC2 together explaining 81.1% of variance for 16S and 34.0% for ITS. 1

The history effect was also reflected in the relative distance structure: history/Round-2 distance ratios were 1.15 for Round-2 lignin and 1.59 for Round-2 lignin plus labile carbon, indicating stronger historical than current-condition effects in these comparisons. 1 In the 16S convergence test, under identical Round-2 conditions, the mean Bray–Curtis distance was 0.507 for L-L versus LC-L and 0.486 for L-LC versus LC-LC, compared with 0.441 for L-L versus L-LC and 0.306 for LC-L versus LC-LC. 1 The reported bacterial 16S memory index was ~0.50, meaning that communities with different histories retained approximately half of the maximum possible divergence under the same current condition. This estimate does not describe fungal ITS communities. 1

Taxonomic composition showed the same historical structuring. Lignin-history groups L-L and L-LC were Pseudomonas-dominated at 52–53%, with low Acinetobacter (<1–20%), whereas lignin-plus-labile-history groups LC-L and LC-LC were Acinetobacter-dominant at 31–34%, with Pseudomonas at 24–26% and Enterobacter at 2.7–18.1%. 1 OTU retention was 81% from L to L-L and 91% from LC to LC-LC, indicating that core taxa were maintained across the passage series. 1

With n=3 per group, no individual OTUs reached FDR (false discovery rate, a multiple-testing correction) significance; the minimum p was 0.10 for Mann–Whitney tests with 3 vs 3. The report therefore gives effect sizes as descriptive CLR (centered log-ratio) differences rather than as statistically significant enrichment. In the 16S comparison of L versus LC, Aeromonas (+6.4) and Shewanella (+6.6) were higher in LC, while Peredibacter (-7.3) and Comamonas (-6.8) were higher in L. The report annotates Aeromonas as copiotrophic and fast-growing, Shewanella as a facultative anaerobe, Peredibacter as a bacterial predator and Comamonas as an aromatic compound degrader. 1

The report’s supporting figures include 16S_ecological_memory_pcoa.png, a PCoA colored by Round-1 history with convex hulls. They also include ecological_memory_distances.png, which compares history-associated and Round-2 condition-associated community distances. 1

Interpretation

These results support the interpretation that an initial carbon regime can alter which taxa persist or become dominant, thereby constraining later community responses when the environment changes. 1 The report interprets the persistent segregation of Round-2 communities by Round-1 history as historical contingency, with first-round history exceeding the effect of current carbon. 1 The report also links the ~0.50 memory index and the observed non-convergence to priority effects. Under that hypothesis, a copiotroph-enriched community established by labile carbon in Round 1 resists convergence with lignin-only histories. The hypothesis is consistent with ecological theory, but the mechanism was not directly measured. 1 The stronger effect of Round-1 history than of current carbon source supports condition-specific-fitness, because the observed legacy depended on the earlier selective regime as well as on the present one. 1

The initial enrichment created a strong ecological filter. The base community was dominated by unclassified taxa (Incertae Sedis, 44.3%), with Pseudomonas at <0.1%. After a single lignin round, Pseudomonas rose to 39.3% and Acinetobacter to 25.2%, together comprising >64% of 16S reads. Over the same round, Shannon diversity declined from 6.46 to 3.16 and observed OTUs declined from 1,594 to 163, a 90% reduction. 1 The report treats this near-complete turnover, together with the Base-versus-Round-1 PERMANOVA (M1; R²=0.992, p=0.008), as support for lignin acting as a strong selective filter. In the report’s reading, the filter favors taxa capable of metabolizing lignin-derived aromatics; that capability is described as consistent with the data rather than measured. 1 The R²=0.979 (p=0.001) value reported alongside the Base-to-L shift is the treatment effect across all 7 groups (M3), not a Base-versus-L test. 1

Adding labile carbon instead increased Acinetobacter to 41.7% (vs 25.2% in L) and Aeromonas to 20.2% (vs 0.1%), while Pseudomonas decreased to 23.0% (vs 39.3%). Shannon diversity was further reduced (2.41 in LC vs 3.16 in L) and Pielou evenness dropped (0.49 vs 0.62). These differences show that the first carbon regime established distinct starting states for subsequent passage. 1 The report interprets these shifts as labile carbon relaxing stringent selection for specialist lignin degraders and enabling fast-growing copiotrophic taxa to co-dominate. It calls the fungal response even more striking: lignin alone selected Fusarium and Fusicolla (soft-rot fungi), while lignin plus labile carbon selected Chrysosporium and Aspergillus, genera more associated with opportunistic saprotrophic growth. 1 It further suggests the hypothesis that labile-carbon priming, a well-established soil-ecology concept, may operate at the community-composition level rather than simply through metabolic activation. On this view, labile carbon restructures the community toward copiotrophic taxa that persist through subsequent passages. These are interpretations of compositional data, not measurements of lignin degradation or metabolic activity. 1

Fungal ITS genera also differed by treatment. Base was Fusarium-dominated (26.5%) with diverse secondary taxa (253 OTUs, Shannon=3.50). In L, Fusarium persisted (43.4%) and Fusicolla emerged (30.3%). LC showed complete replacement, with Chrysosporium dominating (61.4%) alongside Aspergillus (27.2%). In Round 2, Malassezia became dominant in L-L (63.4%) and LC-L (50.0%), while Pleurotus emerged in LC-LC (50.0%). 1 The report notes that the dominance of Aspergillus, Fusarium and Chrysosporium in ITS profiles is consistent with known saprotrophic capabilities of these genera. 1 It calls the Pleurotus emergence in LC-LC biologically notable, because this white-rot basidiomycete is an efficient lignin degrader yet appeared only when labile carbon was available in both rounds. The report explicitly flags the ITS replicate problem when making this point. 1 These genus assignments carry a separate taxonomy caveat. ITS taxonomy was assigned by BLAST against NCBI ITS_RefSeq_Fungi (19,375 reference sequences), some assignments may be imprecise, and genus level is more reliable. The standard UNITE database could not be downloaded, so rare or novel taxa may be missed. All fungal genus patterns also remain subject to the ITS reproducibility caveats below. 1

The bacterial memory result is stronger than the fungal result in this experiment. 1 Fungal Round-2 history was not statistically detectable (R²=0.142, p=0.090), and the overall ITS PERMANOVA R² was lower than for 16S (0.582 vs 0.979). 1 ITS diversity collapsed to 3–8 OTUs per group after enrichment (from 253 in Base). ITS replicate consistency was poor: within-group Bray–Curtis distance was 0.65 and reached 0.99–1.00 for several Round-2 groups, compared with 0.09 for 16S. 1 The report describes bacteria as showing robust, reproducible treatment effects and fungi as showing extreme stochasticity at the same sample size. It suggests that bacteria and fungi differ in their mechanisms of community assembly, but that mechanistic difference is an interpretation, not a demonstrated result. 1 The fungal evidence is therefore preliminary and does not establish a comparable fungal memory effect. 1 This marker-specific contrast motivates integration with multi-omics-integration, while the bacterial result provides a tractable target for linking persistent community composition to gene content through pangenome-integration. 1

The pre-registered per-group Procrustes M² analysis was meant to formally test 16S–ITS concordance, but it was not completed. With within-group Bray–Curtis distances approaching 1.0 for Round-2 groups, rotating near-random ITS points onto the structured 16S ordination would have produced a high M² (poor fit) regardless of biological concordance. Cross-kingdom concordance therefore remains untested rather than refuted at this sample size and sequencing depth. 1

Scope and Limitations

The bacterial history effect was detected in a design with n=3 per group, while the Round-2 factorial analysis used n=12; the smaller per-group replication limits pairwise inference even though the factorial history effect was significant. 1 Pairwise Mann–Whitney U tests had a minimum achievable p-value of 0.10 because there were only 10 possible permutations for 3-versus-3 comparisons, and pairwise PERMANOVA had a permutation floor of approximately p~0.10. 1 The report identifies this n=3 permutation floor as its primary statistical limitation and recommends n>=5 per group to enable pairwise significance testing. 1

PERMDISP, a test of differences in within-group dispersion, was significant for 16S (p=0.0004) and ITS (p=0.0001), so the PERMANOVA signals reflect both differences in group location and differences in dispersion. 1 The experiment used 97% vsearch OTUs rather than amplicon sequence variants, so closely related organisms may have been merged. 1 The reported memory effect is consequently a strong bacterial community-composition result, but its taxonomic and mechanistic basis requires further testing. 1

Open Directions

  • Increase replication to n>=5 per group and repeat the Round-1/Round-2 factorial design to test whether the history effect and the ~0.50 memory index remain stable. 1
  • Use DADA2 amplicon sequence variants, phylogenetic diversity, and UniFrac distances on the existing 16S data to determine whether ecological memory persists beyond 97% OTU definitions. 1
  • Cross-reference persistent Pseudomonas, Acinetobacter, and Comamonas lineages with kbase_ke_pangenome and test whether retained gene-content differences explain the L-versus-LC legacy. 1
  • Infer or measure beta-ketoadipate and protocatechuate pathway content in the retained bacterial taxa to test whether functional lignin-associated capacity predicts persistence across passages. 1
  • Add intermediate passage time points and fit community trajectories to distinguish rapid restructuring from gradual historical persistence. 1
  • Increase ITS sequencing depth, use UNITE-based taxonomy, and repeat the history experiment to test whether the absent fungal memory signal reflects biology or poor replicate structure. 1

Projects cited

  1. Lignin Enrichment and Ecological Memory in Microbial Communitieslignin_community_enrichment · cited 53× · primary source