henhaobio
Sign in

Scientific ReportsEcologyPeer-reviewed journal article

Soil communities following clearcut and salvage harvest have different early successional dynamics compared with post-wildfire patterns

Teresita M. PorteriD, Dave MorrisiD, Emily E. SmenderovaciD, Erik J. S. EmilsoniD, Lisa VenieriD

PDFDOIPreprintCode
In-depth full-text digest available · Jump to digest ↓

Abstract

Understanding the impacts of harvest and subsequent silviculture practices at stand scales on the below-ground biota, and their associated nutrient cycling processes, is needed to more fully evaluate the sustainable management of boreal forest systems. While stand replacing wildfire is the primary natural disturbance mechanism in jack pine-dominated boreal forest systems; clearcut harvest also results in stand renewal so is sometimes used in silvicultural systems to emulate natural disturbance and renewal processes. In this study, we simultaneously assessed the successional trajectories of three major taxa of the below ground soil community, bacteria, fungi, and arthropods using DNA metabarcoding. The objectives of this study were to use a chronosequence framework to: 1) assess whether the soil communities following clearcut harvest and wildfire converge along a successional gradient, 2) assess when the soil community recovers following clearcut harvest to the pre-disturbance, mature, wildfire reference condition, and 3) assess the effects of cumulative disturbance on soil community succession (i.e., wildfire followed by salvage harvesting of fire-killed trees). We found that richness (alpha diversity) did not illustrate any clear patterns of convergence and could, therefore, underestimate recovery times, especially for soil arthropods. Comparisons of the underlying community composition (beta diversity) proved to be more informative. In this case, we found that different soil taxa following clearcut harvest recovered on different timelines compared with succession following stand-replacing wildfire. In general, bacteria appear to be the first to converge to post-wildfire conditions followed by arthropods, however, fungi did not converge within the time frame of the chronosequence. This suggests that more extended periods are required to achieve complete recovery of the soil fungal community to the pre-disturbance condition. The cumulative disturbance associated with salvage harvest appeared to have a greater (compounded) effect on soil communities when compared with wildfire or clearcut harvest. This work showcases the performance of a scalable method for monitoring a diverse arrange of soil biota using DNA metabarcoding. In future work, tracking fungal and arthropod soil communities may provide more insights into the longer-term effects of current forest management practices and provide guidance when comparing alternative approaches. Open Research StatementSequences have been deposited to the NCBI SRA under the GRDI-Ecobiomics project accession PRJNA565010 for the BioSample accessions SAMN26926703 - SAMN26926795 used in this study. The MetaWorks v1.9.3 multi-marker metabarcode bioinformatic pipeline is available from https://github.com/terrimporter/MetaWorks. The ITS classifier based on the ITS UNITE+INSD full dataset v8.2 and trained to work with the RDP Classifier (https://github.com/terrimporter/UNITE_ITSClassifier). The COI classifier v4 is available from https://github.com/terrimporter/CO1Classifier. The code used to produce figures, including infile and metadata files, will be available on https://github.com/terrimporter/Chronosequence_HarvestType.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (11 Nov 2024) · CC BY 4.0

Across 31 jack pine sites, clearcut soil bacteria converge with post-wildfire communities by crown closure, fungi do not, and salvage-logged communities stay distinct for ~20 years.

Boreal forest policy assumes that harvesting which emulates natural disturbance lets stands converge with, and recover toward, post-fire conditions, but below-ground biota are rarely tested because they are hard to survey. Building on an earlier post-wildfire soil succession reference, this work uses metabarcoding of bacteria, fungi and arthropods along 60-year clearcut and 20-year salvage chronosequences to test that assumption.

Key results

  1. SuggestedWhen clearcut and wildfire stands are compared at matched stand development stages, soil bacterial beta diversity converges by crown closure and arthropods appear to converge at early self-thinning (the Discussion says late). Fungal composition stays distinct from establishment through late self-thinning.

    Holm-corrected pairwise PERMANOVA on Sorensen distances: fungal clearcut vs wildfire communities differed at every stage (p-values < 0.05); NMDS stress 0.08 (bacteria), 0.11 (fungi), 0.13 (arthropods). At establishment, wildfire communities tracked higher pH; clearcut ones tracked higher total organic carbon (and total nitrogen for fungi, arthropods). · Assessing convergence and recovery between…; Convergence and recovery of wildfire- and…

  2. SuggestedSoil communities in post-wildfire salvage-logged stands clustered apart from both wildfire and clearcut stands through crown closure (about 20 years after disturbance), and this separation was associated with lower soil total nitrogen.

    NMDS across wildfire, clearcut and salvage sites (stress 0.07 bacteria, 0.11 fungi, 0.13 arthropods); bacterial separation backed by pairwise PERMANOVA; stage-by-disturbance interaction p = 0.001 for all taxa; authors attribute the significant fungal and arthropod wildfire vs clearcut differences at crown closure mainly to unequal dispersion (p-value = 0.004 and 0.00000008). · Comparison of full tree clearcut and salvage…; Impact of salvage logging relative to wildfire…

  3. SuggestedRichness alone pointed to quick recovery of clearcut soil communities toward the mature wildfire condition. Beta diversity instead showed fungal and arthropod communities still distinct from that condition at every sampled stage, although moving toward it.

    Richness: bacteria similar to mature reference at every stage (adjusted p-values >= 0.05), arthropods recovered by crown closure, fungi by late self-thinning. Beta diversity (NMDS stress 0.08, 0.12, 0.14 for bacteria, fungi, arthropods): bacteria differed at establishment and late self-thinning; authors flag unequal dispersion as a possible PERMANOVA confounder. · Assessing convergence and recovery between…; Management Implications

  4. ShownAcross clearcut and wildfire stands, stand development stage explained more community variation than its interaction with disturbance type, and disturbance type alone explained the least.

    PERMANOVA variance explained (bacteria, fungi, arthropods): stand development stage 28%, 22%, 20%; stage-by-disturbance interaction 16%, 12%, 11%; disturbance type alone 4%, 5%, 4% (p-value = 0.001). · Assessing convergence and recovery between…

  5. ShownAt establishment, harvested sites had higher sequence variant richness than wildfire sites, with fungal richness highest after salvage. By crown closure, richness in all three groups matched the wildfire reference.

    At establishment, bacterial and arthropod richness did not differ between clearcut and salvage (p-value > 0.05) but both exceeded wildfire (p-value < 0.05); fungal richness rose in clearcut and further in salvage (p-value < 0.05); at crown closure no group differed from wildfire (p-value > 0.05). · Comparison of full tree clearcut and salvage…; Convergence and recovery of wildfire- and…

Methods, data and reproducibility

Space-for-time chronosequence of jack pine stands grouped by disturbance type and stand development stage. Soil 16S, ITS2 and COI metabarcodes were denoised to sequence variants (SVs), then compared using richness tests, Sorensen-based non-metric multidimensional scaling (NMDS) with envfit, and Holm-corrected pairwise permutational ANOVA (PERMANOVA).

  • Bulk soil (bryophyte, litter-fermentation-humus (LFH) and upper mineral layers, pooled per sample) from jack pine-dominated wildfire, clearcut and salvage stands in Ontario, sampled 2017 · soil bacteria, fungi and arthropods (environmental DNA) · n = 31 sites (14 wildfire, 11 clearcut, 6 salvage); 3 replicates/site · PRJNA565010; BioSamples SAMN26926703 - SAMN26926795
  • Bacterial 16S v4-v5 amplicons (515F-Y/926R primers), Illumina MiSeq 2 x 250 bp paired-end · bacteria (soil community) · PRJNA565010
  • Fungal ITS2 amplicons (ITS9/ITS4 primers), Illumina MiSeq paired-end · fungi (soil community) · PRJNA565010
  • Invertebrate COI amplicons from two markers: F230R_modN (LCO1490/230R_modN) and BE (B/E primers) · arthropods (soil invertebrate community) · PRJNA565010
  • Combined metabarcode reads: 10,077,185 retained after QC in 114,968 SVs, then rare SVs (<0.01% of reads) removed and soil layers pooled · n = 49,721,154 raw pairs; 1,671 SVs (6,642,061 reads) after filtering
  • LFH and mineral horizon (Ae, Bf) samples assayed for pH, total organic carbon and total nitrogen (supplementary Table S1)
Methods & tools
MetaWorks v1.9.3, SEQPREP v1.3.2, CUTADAPT v3.2, VSEARCH v2.15.2, RDP Classifier v2.13, UNITE v8.2-based ITS classifier, COI classifier v4, vegan 2.6-2 (metaMDS, adonis2, betadisper, envfit), pairwiseAdonis 0.4, goeveg 0.5.1, MoBio PowerSoil DNA Isolation Kit, Illumina MiSeq
Software & environment
MetaWorks v1.9.3; SEQPREP v1.3.2; CUTADAPT v3.2; VSEARCH v2.15.2; RDP Classifier v2.13 (UNITE v8.2, COI classifier v4); R v4.1.1 in RStudio 2021.09.0; vegan 2.6-2; goeveg 0.5.1; pairwiseAdonis 0.4; ggpubr 0.4.0; ggplot2 3.4.4; ggmap 4.0.0
Assay conditions & compute
Read processing ran on the General Purpose Science Cluster at Shared Services Canada; no hardware, runtime or memory figures reported.
Reproducibility gaps
Figure code, infiles and metadata are promised ('will be available') at the Chronosequence_HarvestType repo; raw reads are given only as SRA accessions; soil chemistry appears only in supplementary Table S1.
Limitations (5)
  • Beta-diversity dispersion was not homogeneous, so PERMANOVA differences, especially for fungi and arthropods, may partly reflect dispersion rather than true compositional shifts (crown-closure wildfire vs clearcut dispersion p-value = 0.004 fungi, 0.00000008 arthropods). · stated by the authors
  • No harvest-origin stand exceeded 66 years; clearcut sites span only establishment to late self-thinning and salvage sites only establishment to crown closure, so the timing of full recovery cannot be confirmed. · stated by the authors
  • Sparse fine-level taxonomic coverage in reference databases restricted analyses to sequence variants, and functional consequences of the compositional shifts (e.g., nutrient cycling) were not measured. · stated by the authors
  • The space-for-time design assumes time since disturbance is the main difference among sites, yet some stands come from the same large fires and only 6 salvage sites are compared, which limits independence and statistical power. · noted in this summary
  • The text is internally inconsistent. Results place arthropod beta-diversity convergence at early self-thinning, but Discussion puts it at late self-thinning. The establishment stage is 0-15 years in Methods but 0-10 years in Discussion. · noted in this summary

Read this if you study how clearcutting or post-fire salvage logging reshapes soil bacterial, fungal and arthropod succession, or if you evaluate natural disturbance emulation in boreal forestry using DNA metabarcoding.

Citations
6
FWCI
3.37
Field percentile
top 6%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISPorter '22Su '20Whitman '22Woolet '21Whitman '25Johnson '26
86% semantic matchEcology
T. M. Porter, E. Smenderovac, D. Morris+1 more·bioRxiv·Nov 18, 2022