Corpus 1,272 assessed · 1,173 scored · 643 reproduced ≥75 · 168 flagged ·∅ 74.1/100
← New search

Spatial patterns of benthic biofilm diversity among streams draining proglacial floodplains.

Front Microbiol · 2022
95/100 3/4
Why this verdict

The main results reproduced: recomputed values matched the published ones within tolerance.

Reproduced on the brainbox compute brainarbeit.com
How its reproducibility compares
95/100
Reproducibility score
1.2 SD above mean
vs. all fields · 1173 studies
🎯 Scores higher than 89% of all assessed papers rank 105 of 1173 scored

A 0–100 reproducibility-quality score from the per-question grades, shown as a z-score: standard deviations above (+) or below (−) the mean of comparable assessments.

Reproduction agent’s raw note

REPRODUCED (1:1 on downstream pipeline results). The 6 authored R scripts + Zenodo 6424496 processed tables deterministically regenerate the paper's reported diversity/network/biomass statistics. Ran R 4.3.3 (vegan/igraph) on «our HPC» mirroring the authors' code. 33 graded claims: 23 EXACT (incl. ASV total 53858, all four PERMANOVA R2 9.3/9.7/2.1/0.6%, every network Table-3 metric edges/nodes/diameter/mean-distance/composition, specific-BCP ANOVA F=14.530 to 3 decimals, Proteobacteria/Planctomycetota ASV counts), 9 WITHIN-TOLERANCE (stochastic NMDS/louvain, off-by-one taxonomy-rank counts from blank-bin convention, biomass fold ratios, weighting-convention Venn), 0 MISMATCH, 1 NOT-ATTEMPTED (Procrustes 0.43 - genuinely absent from the shipped code). No fabrication indicators: the author-entered Fig-2 Venn proportions are independently reproducible from the raw 16S table to +/-1pp. Out of scope: wet-lab assays and the upstream QIIME2 re-run from raw reads. The paper is described well enough to reproduce and its shipped artifacts deliver what is promised.

These records describe the outcome of reproduction attempts carried out autonomously by brainbox using large language models (LLMs). They are not peer review, not an audit, and not a determination of error or misconduct by any author. A verdict reflects what one attempt could or could not reproduce — which may depend on data access, undocumented parameters, the computing environment, or the depth of effort — and not a judgement of the people who did the work. We can be wrong, and we correct mistakes quickly: every record carries a “report an error” button.

✎ I am an author of this paper

Updated or fixed a deposit, or is there an erratum? Ask us to re-run the metrics. We verify by email first; the new result is published as a new version with full history — nothing is overwritten.

Reason for the rerun

We email you a confirmation link first. The rerun is an objective re-measurement — it cannot change the verdict in your favour, only ask us to look again.

Provenance — full disclosure

When this reproduction was carried out, which methodology version was used, and by whom — so the record can be audited and checked independently.

Reproduced
2026-06-26
Rubric version
not recorded
Assessed by
Last updated
2026-08-05

Provisional, curator- or AI-assessed, and independently checkable. A reproduction outcome states what one attempt could reproduce — not a judgement of the authors.

Deep full-text extraction

Model: sonnet
Founding hypothesis

The authors hypothesized that bacterial diversity and community structure differ between braided glacier-fed streams and their lateral tributaries (a lateral chronosequence), and that aquatic primary producers (e.g., algae) shape the bacterial communities of benthic biofilms in proglacial floodplains.

Core claims
  • Benthic biofilms in tributaries develop higher biomass than those in glacier-fed streams along the lateral chronosequence finding
  • Diversity and community composition of benthic biofilms differ markedly between glacier-fed streams and tributaries finding
  • Bacterial communities in glacier-fed streams show spatial turnover along the longitudinal chronosequence finding
  • Environmental filtering is an underlying mechanism producing these unexpected spatial patterns despite close proximity/connectivity of streams mechanism
  • Photoautotrophic communities shape bacterial communities across the various proglacial streams finding
  • Physicochemical characteristics (temperature, conductivity, turbidity, DOC, DIN) markedly differ between glacier-fed streams and tributaries across all three floodplains finding
  • 25 bacterial genera were identified as enriched in ASVs with high likelihood of being exclusive to glacier-fed streams, indicating risk of loss as glaciers shrink finding
Experimental setups
Assay System Perturbation Readout Platform
16S rRNA gene amplicon sequencing (V3-V4, 341F/785R primers) benthic sediment biofilms, glacier-fed and tributary streams, 3 Swiss Alps floodplains none (comparative, stream type/season/floodplain) bacterial ASV community composition, richness, diversity Illumina MiSeq, QIIME2, SILVA v138.1
18S rRNA gene amplicon sequencing (V4, TAReuk454F/TAReukREV3 primers) benthic sediment biofilms, glacier-fed and tributary streams none (comparative) eukaryotic phototroph OTU community composition, richness Illumina MiSeq, QIIME2/vsearch, SILVA v138.1
Flow cytometry benthic sediment (detached cells) none bacterial abundance (cells per gram dry sediment) NovoCyte (ACEA Biosciences), SybrGreen staining
Ethanol extraction chlorophyll-a assay benthic sediment none chlorophyll-a concentration (proxy for phototroph biomass) plate reader (436/680 nm)
EPS extraction (EDTA extraction, glucose equivalent assay) lyophilized benthic sediment none extracellular polymeric substance concentration
[3H]-leucine incorporation assay benthic sediment none bacterial carbon production liquid scintillation counter (Tri-Carb 4910 TR, Perkin Elmer)
Streamwater chemistry (DOC, inorganic nutrients, major ions) streamwater, glacier-fed and tributary streams none DOC, ammonium, nitrite, nitrate, soluble reactive phosphorus concentrations Sievers M5310c TOC Analyzer; LaChat QuikChem 8500
In situ physicochemical monitoring streamwater none temperature, pH, dissolved oxygen, conductivity, turbidity WTW Multi 3630 IDS, WTW TetraCon 925, PME Cyclops-7
Key results
  • Streamwater temperature, DOC concentration, and electrical conductivity were significantly higher in tributaries than glacier-fed streams
  • Turbidity was significantly higher in glacier-fed streams and negligible in tributaries
  • Dissolved inorganic nitrogen (DIN) concentration was higher in glacier-fed streams than tributaries
  • PCA showed physicochemical characteristics markedly differed between GFS and TRIB in all three floodplains
  • 16S rRNA sequencing yielded 257 amplicon libraries and 32,186,859 total reads (avg. 124,273 reads/sample)
  • 18S rRNA sequencing yielded 242 amplicon libraries and 21,413,150 total reads, clustered into a 429-OTU phototroph dataset
  • 25 bacterial genera identified as significantly enriched among ASVs likely exclusive to glacier-fed streams
Key statistics
  • pvalue p < 0.001 (ANOVA for temperature, DOC, conductivity, turbidity differences between GFS and TRIB)
  • pvalue p < 0.001 (ANOVA for DIN concentration difference between GFS and TRIB)
  • count 259 (total benthic sediment samples collected across three floodplains)
  • count 257 libraries; 32,186,859 reads (16S rRNA amplicon sequencing dataset (4 samples discarded))
  • mean 124,273 reads/sample (average 16S rRNA sequencing depth per sample)
  • count 242 libraries; 21,413,150 reads (18S rRNA amplicon sequencing dataset (19 samples discarded))
  • count 429 OTUs (18S rRNA phototroph dataset after filtering singletons/non-phototrophs)
  • other 4.93 ± 3.84% (coefficient of variation among flow cytometry technical replicates)

Statistical methods review

Model: sonnet

A neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.

The study compared benthic biofilm properties and prokaryotic/eukaryotic (16S/18S rRNA amplicon) community diversity between glacier-fed streams (GFS) and tributary streams (TRIB) across three proglacial floodplains and two seasons. Univariate biomass indicators (bacterial abundance, chlorophyll-a, EPS, bacterial carbon production) were analyzed with three-way ANOVAs and t-tests, while community-level data were analyzed with PERMANOVA, NMDS, betadisper, procrustes, breakaway/betta alpha-diversity estimation, and Fisher's exact tests for ASV-stream type associations. Multiple-comparison p-values were adjusted using the Benjamini-Hochberg method, and prokaryote-eukaryote co-occurrence networks were built by averaging SparCC, Spearman correlation, and SpiecEasi association matrices.

Replicationmixed Sample sizeSample sizes are given descriptively per floodplain and stream type (Table 1: 259 total sediment samples; 257 16S libraries after QC; 242 18S libraries; 198 samples used for co-occurrence networks, 92 GFS/106 TRIB); technical replicates (n=3) were used for flow cytometry bacterial abundance measurements. No formal power analysis or a priori sample size justification is described. GroupsGlacier-fed streams (GFS) vs. tributary streams (TRIB), across three floodplains (Otemma, Val Roseg, Valsorey) and two seasons (early, late) Pairingunclear Randomization/blindingstated DispersionSD Exact p-valuesno Multiplicity correctionBenjamini-Hochberg (BH) false discovery rate correction
Statistical tests used
Test Applied to n Assumptions
Three-way ANOVA Differences in bacterial abundance, chlorophyll-a, EPS, and bacterial carbon production across floodplain, stream type, and season not stated
t-test (BH-adjusted) Stream type (GFS vs. TRIB) differences for each biomass indicator within each floodplain not stated
PERMANOVA (adonis, Euclidean distance) Effects of floodplain, stream type, and season on streamwater physicochemical characteristics 40 water sampling sites (per Figure 1 legend) not stated
PERMANOVA (adonis, Bray-Curtis dissimilarity) Effects of stream type, floodplain, season, and chlorophyll-a on 16S bacterial community composition not stated
pairwise.adonis (BH-adjusted) Pairwise comparisons of community similarity between stream types within each floodplain not stated
betadisper / t-test Multivariate homogeneity of group dispersions (compositional variability) between stream types on Bray-Curtis matrices not stated
breakaway / betta 16S ASV and 18S OTU richness estimation and significance testing between stream types not stated
Fisher's exact test (odds ratio, BH-adjusted) Likelihood of individual ASVs being exclusively present in glacier-fed streams vs. tributaries 25 genera retained after adjustment not stated
Approaches that could also have been used
  • Stream-type differences in each biomass indicator (bacterial abundance, chlorophyll-a, EPS, BCP) were assessed with multiple t-tests per floodplain, with p-values then adjusted using the Benjamini-Hochberg method.
    Could also: A mixed-effects ANOVA or the existing three-way ANOVA framework extended with a post-hoc test (e.g., Tukey HSD or estimated marginal means contrasts) — This would test stream type, floodplain, and season effects (and their interactions) within a single model, which can offer a unified family-wise error control and directly estimate interaction effects rather than relying on separate pairwise tests per floodplain.
  • Community composition differences were visualized with NMDS on Bray-Curtis dissimilarities.
    Could also: Principal Coordinates Analysis (PCoA) on the same distance matrix — PCoA produces axes with an explicit proportion of variance explained, which can complement NMDS's rank-based stress-minimization approach when quantifying how much variation each ordination axis captures.
  • Differences in bacterial community composition were tested with PERMANOVA (adonis) on Bray-Curtis dissimilarities across stream type, floodplain, and season.
    Could also: A distance-based redundancy analysis (db-RDA) or a Mantel/partial Mantel test — These approaches can incorporate continuous environmental covariates (e.g., chlorophyll-a, temperature) alongside categorical factors within the same constrained ordination framework, which can help partition variance explained by multiple continuous and categorical predictors simultaneously.
  • Alpha diversity (richness) was estimated using the breakaway estimator, with significance testing via betta.
    Could also: Rarefaction-based diversity indices (e.g., Shannon, Simpson) or Hill-number-based diversity estimation (e.g., iNEXT) — These provide complementary diversity metrics that weight abundant vs. rare taxa differently, which can be reported alongside richness estimates to give a fuller picture of diversity patterns across samples of uneven sequencing depth.
  • Association between individual ASVs and stream type (GFS-exclusive presence) was tested using Fisher's exact test on 2x2 contingency tables, with BH-adjusted p-values.
    Could also: A generalized linear model (e.g., logistic regression) with stream type, floodplain, and season as covariates — A GLM framework would allow simultaneous adjustment for floodplain and season while estimating the stream-type association for each ASV, rather than testing stream type in isolation within pairwise tables.
  • Co-occurrence networks were built by averaging distance matrices from SparCC, Spearman correlation, and SpiecEasi (mb method), retaining the top 10% of edges by interaction strength.
    Could also: Reporting network edge stability via bootstrapping or permutation-based null models for each individual method — This would provide a confidence measure for retained edges and allow comparison of how consistently each method recovers the same associations, complementing the consensus-averaging approach already used.
Software: R 4.0.3 · vegan (adonis, betadisper, procrustes, ordisurf) · factoextra (fviz_pca_biplot) · breakaway / betta · QIIME2 · pairwise.adonis · SpiecEasi (Meinshausen and Bühlmann method) · igraph

What was reproduced

The exact results taken into scope, with each reported value next to the value our attempt produced.

Scope — pmid-36003939

Paper: Brandani et al. 2022, "Spatial patterns of benthic biofilm diversity among streams draining proglacial floodplains." Front Microbiol 13:948165. DOI 10.3389/fmicb.2022.948165 · PMCID PMC9393633

Artifacts

  • Analysis code: https://github.com/jadebrandani/Proglacial_floodplain_diversity — 6 R scripts (R 4.0.3). NOT the upstream amplicon pipeline; these are the downstream diversity/stats scripts that consume processed tables.
  • Processed data (feeds the R scripts): Zenodo 10.5281/zenodo.6424496 (data.zip, ~7.4 MB) — 16S_table.csv, 16S_Taxonomy.csv, 16S_Metadata.csv, 18S_otu_table.csv, 18S_Taxonomy_NEW.csv, 18S_Metadata.csv, water_data.csv, Biomass_data.csv, network subfolder.
  • Raw reads: SRA PRJNA808857 (16S V3-V4 341F/785R; 18S V4 TAReuk454F-TAReukREV3).

Pipeline map (which result comes from which pipeline)

Reported result Pipeline / script In scope?
Raw read processing → ASV table (QIIME2 + Trimmomatic, SILVA v138.1) QIIME2 (authors' external run from SRA) partial — upstream; very heavy. Attempt only if time allows after core.
Total 16S ASV count (53,858) derived from 16S_table.csv rows yes (third-party-equivalent: count on shipped table)
16S taxonomy counts (55 phyla … 1,061 genera; per-phylum ASV counts) 16S_Taxonomy.csv summarisation yes
# samples / libraries (257 16S, 242 18S, 259 sediment) table column / metadata counts yes
Alpha diversity richness (TRIB 2,461±626 vs GFS 1,921±509, breakaway) Alpha_diversity.R (breakaway/betta) yes
Pielou evenness (no diff GFS vs TRIB) Alpha_diversity.R yes
18S OTU counts (3,553 total; 429 phototroph) 18S_otu_table.csv / taxonomy yes
Phototrophs = 62.2% eukaryotic rel. abundance 18S table summarisation yes
Beta diversity: NMDS stress (16S 0.19, 18S 0.24); PERMANOVA; betadisper; procrustes 0.43 Betadiversity_16S_18S.R (vegan) yes
Venn shared/unique ASV proportions (Fig 2) Alpha_diversity.R / MicEco ps_venn yes
Fisher enrichment / odds ratios Fischer_enrichment.R yes
Water physicochemistry PCA + ANOVA Water_physiochemical.R yes
Biomass: Chl-a 35× higher TRIB, bact. abundance 10× greater, BCP, EPS (Table 2) Sediment_biomass.R yes
Co-occurrence network stats (Table 3: nodes/edges/clusters/modularity) Network_Analysis.R (SparCC/SpiecEasi consensus + igraph) yes (final stats); consensus inference is stochastic/heavy — network CSVs are shipped, so recompute graph metrics from those

Out of scope (not attempted)

  • Wet-lab measurements (the generation of Chl-a, cell counts, BCP, EPS, water ions in the lab) — we reproduce the statistics over those numbers, not the assays.
  • DNA extraction / sequencing.
  • SparCC/SpiecEasi de-novo network inference from scratch is stochastic and heavy; the repo ships the resulting edge lists, so we recompute the reported graph metrics (igraph) from the shipped network files rather than re-inferring edges.

Strategy

Run the core computations of the 6 R scripts on the Zenodo data on «our HPC» (conda R env built INSIDE the SLURM job, data on «infra»). Compare each reproduced number to the reported value in claims.tsv. Core ~80% = alpha/beta/taxonomy/biomass counts (cheap, fast); then push into networks + Venn + Fisher. Profile both Zenodo + SRA datasets.

Refinements after reading code (2026-06-26)

  • Commit pinned & confirmed: 1abbd6ae46c0d66e016e8b921e1a2a915c219c89 (2022-04-12).
  • Sample-removal steps (must mirror exactly): drop columns VAR_61,VAR_62 from 16S_table.csv (259→257); drop metadata rows OTE_27,OTE_48,VAR_61,VAR_62.
  • 18S photoautotroph filter (Alpha/Beta scripts): subset_taxa(D_2=="Chloroplastida" | D_3=="Ochrophyta" | D_2=="Cryptomonadales") → prune taxa_sums>1 → keep taxa in >1 sample → keep sampl
Figures / tables: TableFig 3Fig 2
C1
Reported
53858
Reproduced
53858
exact
C2
Reported
257
Reproduced
257
exact
C3
Reported
429
Reproduced
429
exact
C4a
Reported
2461+/-626
Reproduced
2450+/-633
within tolerance
C4b
Reported
1921+/-509
Reproduced
1916+/-515
within tolerance
C5
Reported
55/162/402/599/1061 phyla/class/order/fam/genus
Reproduced
55/161/401/598/1060
within tolerance
C6a
Reported
13503
Reproduced
13503
exact
C6b
Reported
6771
Reproduced
6721
within tolerance
C6c
Reported
5977
Reproduced
5977
exact
C7
Reported
PERMANOVA R2 9.3/9.7/2.1/0.6%
Reproduced
9.33/9.74/2.07/0.65%
exact
C8a
Reported
0.19
Reproduced
0.181
within tolerance
C8b
Reported
0.24
Reproduced
0.236
within tolerance
C9
Reported
GFS 165/113/2.81/3.36
Reproduced
165/113/2.81/3.36
exact
C10
Reported
TRIB 205/167/9.83/9.59
Reproduced
205/167/9.83/9.59
exact
C11
Reported
GFS 69/18/12%
Reproduced
69.0/18.6/12.4%
exact
C12
Reported
TRIB 64.7/29.3/6.0%
Reproduced
64.7/29.3/6.0%
exact
C13
Reported
Venn unique-TRIB 34-47%, shared-W 72-93%
Reproduced
matches shipped Venn CSV +/-1pp (recomputed from 16S_table)
exact
C14
Reported
Chla 35x / bact 10x
Reproduced
35.5x / 9.8x
exact
C15
Reported
14.530
Reproduced
14.530
exact
Cproc
Reported
Procrustes 0.43
Reproduced
NOT-ATTEMPTED (not in repo code)
m.public.grade.not-attempted

Assessments & scoring basis

Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.

No assessment has been recorded yet.
🤝
Reproduced automatically — and fairly

Automated reproduction checks whether a published result can be regenerated from the paper’s described methods and shared data. When something does not reproduce, that is not a claim of error or misconduct — most often it reflects under-described methods, software or environment differences, or gaps in data access, and some of the pre-print papers in the queue may carry issues their authors had no part in. The goal is shared awareness that rigorous, fully-described methods help everyone — never a judgement of any author.

Are you an author? We would genuinely like to hear from you — to clarify the record, add data or code, re-run the pipeline after an accession update, and publish your response right next to the assessment. Everything here is open and auditable.

🚩 Report an error in this record

Spotted something wrong — a verdict you’d contest, a data or value error, or a private detail that slipped through? Tell us, with a short justification. Authors and readers are equally welcome to write in; we review every report.

Prefer email, or the form below not working? Contact us at support@doesitreproduce.com.